Rescuing the World Bank:
A CGD Working Group Report
and Selected Essays
Preface
vii
A CGD Working Group Report
The Hardest Job in the World:
Five Crucial Tasks for the
New President of the World Bank
13
Prepared by a Center for
Global Development Working Group,
Nancy Birdsall and Devesh Kapur, Co-Chairs
Selected Essays
A Global Credit Club,
Not Another Development Agency
69
by Nancy Birdsall
Votes and Voice: Reforming Governance
at the World Bank
87
by Masood Ahmed
The Battle for the Bank
by Ngaire Woods
95
The World Bank and Low-Income Countries:
The Escalating Agenda
103
by William Easterly
The Role of the Bank
in Low-Income Countries
by Steven Radelet
109
Has the World Bank Lost Control?
117
by Adam Lerrick
The World Bank and
the Middle-Income Countries
by David de Ferranti
133
The Missing Mandate: Global Public Goods
by Michael Kremer
The “Knowledge” Bank
by Devesh Kapur
159
Evaluations and Aid Effectiveness
by Pierre Jacquet
The Evaluation Agenda
153
171
183
by Ruth Levine/William D. Savedoff
The World Bank: Buy, Sell or Hold?
by Mark Stoleson
195
v
CGD0502 0527_Engl_6x9.indd 1
8/17/06 2:54:54 PM
CGD0502 0527_Engl_6x9.indd 6
8/17/06 2:54:54 PM
Preface to
Rescuing the World Bank
A CGD Working Group Report
and Selected Essays
T
he development and transformation of societies
where poverty and disease afflict large numbers of
people is a central challenge of the 21st century.
In broad terms at least, that point, once agreed,
seems to justify the existence of the World Bank—still
possibly the single best-placed institution to address
that challenge.
But the World Bank has been under siege—assailed
by critics of the left, right and center on grounds it is not
effective, not accountable, not democratic or legitimate,
and most threatening of all for the Bank, no longer relevant
in a global economy where private capital, production
and ideas dominate.
Yet the world does need a strong World Bank. Without
reform and revitalization at the Bank agreed by its
members—the advanced economies and the poorest—
the world will have one less institution to manage not only
“development” in the conventional sense of the word,
but the related global challenges of the 21st century.
Are the Bank’s current shortcomings exaggerated, or
are they potentially fatal, marking a moment that will later
be seen as the start of a long decline into irrelevance and
obscurity? If they are potentially fatal, can this critical
institution be rescued?
For the most part the contributors to this book see the
Bank as an institution at risk—in a way it has not been
before—but one that can be rescued. Readers will find
here different positions on the what, why and wherefore
of the Bank’s current weaknesses—and on what exactly
ought to be done with what priority to rescue it. These
differences notwithstanding, the book is packed with
practical suggestions that the shareholders (the Bank’s
184 member governments) and the president of the Bank
can take to move their institution toward greater strength,
flexibility, and effectiveness in a rapidly changing global
economic system.
CGD0502 0527_Engl_6x9.indd 7
vii
8/17/06 2:54:54 PM
viii
Rescuing the World Bank
This book has two parts. The first dates to the
appointment of Paul Wolfowitz as the new president
of the Bank in the spring of 2005. We at the Center
for Global Development saw the logic of drawing up
an agenda for him, in support of anticipated efforts to
revitalize the Bank. The Working Group we organized
focused not on internal management issues but on the
structural changes—in mandate, instruments, and its own
governance—that are critical to a renewed Bank. The
group defined five tasks on which the president should
exercise leadership in persuading the Bank’s member
governments to take action. The report of the group,
The Hardest Job in the World: Five Crucial Tasks for the
New President of the World Bank, was made available
to President Wolfowitz on his official entry to the Bank
on June 1, 2005. It is presented in its original form here
as part one of this monograph. To my mind, the tasks
remain as crucial today as they were when the report
was first issued.
The second part has its roots in the September
2005 Annual Meetings of the IMF and the World Bank
in Washington D.C. To generate healthy discussion
and debate and to explore the recommendations with
a broader community, including officials and private
sector actors attending the Annual Meetings, the Center
organized a symposium on the Friday preceding the
weekend meetings. We were fortunate to have President
Wolfowitz himself kick off the symposium by thanking the
Center and the members of the Working Group for their
effort—and noting that if his job was not the “hardest”
in the world it was certainly the “best”. The second
part of this volume comprises a set of essays largely
based on the presentations made that day to a large and
enthusiastic audience.
Readers of the essays will note substantial differences
between some authors on some issues. That was the point
of the symposium and is the point of this publication—to
contribute to the healthy discourse on a set of difficult
issues with no easy answers. In that sense, this volume
represents the kind of contribution the Center for Global
Development strives to make on the many difficult issues
confronting the development community today.
I would like to extend my thanks to three groups of
people who were responsible for bringing this volume
to life.
CGD0502 0527_Engl_6x9.indd 8
8/17/06 2:54:54 PM
ix
A CGD Working Group Report
and Selected Essays
The first is the team responsible for The Hardest Job
in the World. We were fortunate to convince a small
group of 18 distinguished colleagues—from the private
sector, academia, civil society, and the governments of
rich and developing countries—to come together over
the course of three months to map out an agenda for
the new president. I was particularly fortunate to have
Devesh Kapur, a CGD non-resident fellow (now at the
University of Pennsylvania) and co-author of The World
Bank: Its First Half Century, an authoritative two-volume
history of the Bank, join me in chairing the group. His
encyclopedic knowledge of the Bank and his creative
insights were a priceless asset to our work. The work of
the Working Group was coordinated by Milan Vaishnav,
my former special assistant who is now a graduate
student at Columbia. As in every other task, he was
superb, bringing a steady hand and fine judgment to
every step of the process.
In addition, many others who could not participate in
the group also provided input and comments, including
Masood Ahmed, Jessica Einhorn, Ravi Kanbur, Maureen
Lewis, Johannes Linn, Peter McPherson, and John
Sewell. In particular, I would like to thank Kemal Dervis
(former CGD visiting fellow who is now head of the United
Nations Development Programme), John Hicklin (also a
former CGD visiting fellow, while on sabbatical from the
IMF), and David Peretz (former U.K. executive director
at the Bank and IMF) for their extensive comments on
the draft.
Second, I thank the participants of the Annual
Meetings symposium and other authors whose work
appears as essays in the second part of the volume.
Without their willingness to review and revise, and their
patience, we would have missed a unique opportunity
to showcase the thoughtful and provocative analysis
and recommendations of some of the world’s foremost
experts on the World Bank.
Last, but certainly not least, I thank heartily Lawrence
MacDonald and his outstanding communications and
publications staff for their support in the production of
this volume.
CGD0502 0527_Engl_6x9.indd 9
8/17/06 2:54:54 PM
x
Rescuing the World Bank
I hope that this book will contribute to the growing
momentum for change at the World Bank. Change seems
crucial to me and other contributors to this volume. Only
by changing to meet the demands of the 21st century
can the World Bank live up to its potential to be a
powerful, positive force for greater and more broadly
shared prosperity.
Nancy Birdsall
President
Center for Global Development
June 1, 2006
CGD0502 0527_Engl_6x9.indd 10
8/17/06 2:54:55 PM
A CGD WORKING
GROUP REPORT
CGD0502 0527_Engl_6x9.indd 1
8/17/06 2:54:55 PM
CGD0502 0527_Engl_6x9.indd 2
8/17/06 2:54:55 PM
Preface to
The Hardest Job in the World:
Five Crucial Tasks for the
New President of the
World Bank
T
he Center for Global Development has a special
interest in the World Bank. Compared with the
Bank, we are a small institution. But our mission
is virtually the same—to reduce poverty and inequality,
to maximize the benefits of globalization for developing
countries and their poor, and to improve people’s lives.
That is no accident. I am myself a former World Bank staff
member who looks back with pride and satisfaction on
the opportunity the Bank gave me to contribute (modestly,
for sure) to that grand mission, and almost all of my
colleagues here have similar experience and missionary
zeal about the great development project. With the
advantage of complete independence, we do research
and engage actively on how the rich world and the global
institutions—including the World Bank—can better affect
the poor world. At our launch in November 2001, we were
honored by the presence of James Wolfensohn, then in
his sixth year as World Bank president, who framed it
very well indeed, saying, in all good humor, that he hoped
we’d be tough on him and on the Bank.
It was natural then, when we learned in February
of James Wolfensohn’s departure from the Bank, to
begin thinking about the risks and the opportunities
his successor would face. I asked a small group of
distinguished colleagues—from the private sector,
academia, civil society, and the governments of rich
and developing countries—to join a working group to
discuss and make recommendations addressed to the
new president. I was fortunate to persuade Devesh Kapur,
a non-resident fellow of the Center (now at Harvard) and
the co-author of the authoritative book on the World
Bank’s history, that he should join me in chairing the
group. Without his help, his insight, and his patience,
this report would not be what it is.
3
CGD0502 0527_Engl_6x9.indd 3
8/17/06 2:54:55 PM
4
Rescuing the World Bank
Members of the working group met three times. First
in February 2005, shortly after the candidacy of Paul
Wolfowitz was announced. Second in March and finally
in late April. This report would not have been possible
without their willingness to contribute their time, their
energy, and most important their good wisdom and good
judgment to our deliberations. Along with Devesh, I thank
them enormously for their interest and dedication, their
insights, their issue notes, and their continuous stream
of thoughtful comments on working drafts. Many others
who could not participate in the group also provided input
and comments. We especially thank Masood Ahmed,
Jessica Einhorn, Ravi Kanbur, Maureen Lewis, Johannes
Linn, Peter McPherson, and John Sewell.
Our thanks go also to three others. Kemal Dervis, a nonresident fellow of the center, was named the administrator
of the United Nations Development Programme in the
middle of our deliberations. He played a key role in
shaping our deliberations from the inception of the group,
especially on issues of the Bank’s role in global economic
governance. John Hicklin, a visiting fellow at the Center on
leave from the International Monetary Fund, participated in
the group’s discussions and consultations in his personal
capacity and provided thoughtful comments on a range of
issues. The perspective offered by David Peretz, formerly
a senior U.K. Treasury official and executive director of
the Bank and the Fund, was also invaluable.
Finally, my own thanks go above all to my companionin-arms and special assistant Milan Vaishnav. He is at the
beginning of what will surely be a long and successful
career. He brought his own good questions, political
insights, and instinctively good judgment to our project,
as well as critical attention to detail and timeliness. He
was, above all, patient with my bad habit of last minute
changes at late hours. Without colleagues at the Center,
especially Lawrence MacDonald and his superb team, and
Gunilla Pettersson, Milan and I could not have crossed
the finish line.
Readers of our report will see that Working Group
members started from a shared assumption: that the
world needs a strong World Bank. A central challenge
of the 21st century is securing sustainable growth and
poverty reduction in the developing world, where five of
every six people live today (and eight of every nine will live
in less than 50 years). The Bank is perhaps the world’s
CGD0502 0527_Engl_6x9.indd 4
8/17/06 2:54:55 PM
5
A CGD Working Group Report
single best-placed institution to address that challenge.
To do so effectively, the Bank needs to change however,
adapting quickly its mid-20th century policies and habits
to the greatly changed global environment.
Our Working Group focused not on internal management
issues but on the structural changes—in mandate,
instruments, pricing, and its own governance—that are
critical to a revitalized Bank. We look to Mr. Wolfowitz
to take bold leadership in pushing for those changes
through cajoling and consensus-building with the Bank’s
member governments. We look to the Bank’s many
constituents, including civil society groups concerned
with social justice around the world, to support him in
pushing for those changes. The challenge now belongs
to him to exploit the potential of what we call, with good
reason, the hardest job in the world. We hope this report
helps guide him in that challenge.
Nancy Birdsall
President
Center for Global Development
June 1, 2005
CGD0502 0527_Engl_6x9.indd 5
8/17/06 2:54:55 PM
CGD0502 0527_Engl_6x9.indd 6
8/17/06 2:54:55 PM
The Hardest Job in the World:
Five Crucial Tasks for the
New President of the
World Bank
Executive Summary
T
his report sets out five crucial tasks for the World
Bank president to tackle over the next five years.
They are tasks for which the president—through
a combination of charm, cajoling, and horse-trading—
must corral the Bank’s recalcitrant collective of member
governments, including its single largest shareholder, the
United States, to take action—action critical to securing
the Bank’s credibility, legitimacy and effectiveness for
the 21st century.
The five tasks are informed by a set of guiding
principles on which members of the Working Group
agree (see box).
Guiding Principles for the New President
• The Bank’s mission and in-country priorities. The Bank’s
agreed mission (reducing poverty through equitable
growth) provides no real guidance on country-specific
priorities. It is time to end the confusion between what is
good for development in general and what the Bank itself
should do in a particular setting. In today’s complex donor
system, the Bank need not do everything everywhere. It
should take leadership on the idea of partnership with a
country’s own and with other international efforts.
• Equitable growth and political savvy. Rich-country support
for the Bank demands that the Bank’s engagement and
financing in borrowing countries leverage policies be
pro-poor and supportive in general of a more secure and
sustainable global system. But such “leverage” cannot
rely on the detailed conditionality of a “nanny Bank.”
It must rely on Bank staff’s being politically savvy—
sensitive to a country’s political constraints and to the
opportunities of responsible leaders to push reforms.
That implies a premium on systematic analysis of local
politics and institutions—and on increasing Bank-wide
research and analysis of country governance.
7
CGD0502 0527_Engl_6x9.indd 7
8/17/06 2:54:55 PM
8
Rescuing the World Bank
• The Bank as development’s brain trust. The Bank’s
singular comparative advantage is its staff’s broadranging knowledge and experience on the full range of
technical, sectoral and economic issues of development.
The resulting brain trust cannot be unbundled from the
Bank’s financing role, however. It is lending that triggers
and supports policy dialogue and advice to countries,
and it is the income from lending that helps finance the
brain trust.
• The Bank’s governance: toward greater legitimacy and
effectiveness. The Bank’s legitimacy and effectiveness
going forward require that its borrowers be better
represented in its governance. It should undergo a
transformation from a development agency to something
closer in spirit to that of a global “club” in which today’s
developing-country beneficiaries, not only its richcountry benefactors, have a keen sense of ownership
and financial responsibility.
The five priority tasks are:
• Revitalize the World Bank’s role in China, India, and
the middle-income countries.
• Bring new discipline and greater differentiation to
low-income country operations.
• Take leadership on ensuring truly independent
evaluation of the impact of Bank and other aidsupported programs.
• Obtain an explicit mandate, an adequate grant
instrument, and a special governance arrangement
for the Bank’s work on global public goods.
• Push the Bank’s member governments to make the
Bank’s governance more representative and thus
more legitimate.
Revitalize the World Bank’s Role in China,
India, and the Middle-Income Countries
Borrowing from this group of countries has declined
dramatically, because of the high “hassle” costs of dealing
with the Bank and because of their increasing (though
at times uncertain and costly) access to private capital
markets. Their reduced borrowing puts at risk the Bank’s
maintenance of its global expertise, its ability to leverage
equitable and sustainable policies, and its net income over
CGD0502 0527_Engl_6x9.indd 8
8/17/06 2:54:56 PM
9
A CGD Working Group Report
the long run. To remain relevant for these countries, whose
participation in the global club matters for global progress,
the Bank must transform the way it does business. The
new president should:
• Ask the shareholders to review the charter to
determine if the provision that International Bank
for Reconstruction and Development loans be
guaranteed by a sovereign borrower has stifled the
Bank Group’s ability to catalyze private investment,
lend to municipal and other nonsovereign entities,
support deepening of local capital markets, and in
general respond more effectively to the changing
demands of its key borrowers, especially for its more
active and strategic involvement in catalyzing local
and foreign investment.
• Find ways to sharply expand the range of financial
products and instruments now available to borrowers,
such as products and instruments to hedge against
commodity and other risks, better use of the guarantee
function, and, by the Bank itself borrowing in local
currency or in a mix of emerging market currencies,
making it possible for countries to borrow from the
Bank in their own currency.
• Create a new loan product that would visibly reduce
hassle costs for creditworthy countries with reasonably
good performance in economic management and an
adequate record of enforcing environmental and other
safeguards.
• Introduce differential pricing among International Bank
for Reconstruction and Development borrowers, tied
strictly to per capita income (not to credit rating),
to encourage less borrowing for the right reason—
ushering in de facto “graduation” without any arbitrary
rule-based loss of access.
• Explore other pricing or product innovations that
would create incentives for borrowers to make their
own public revenue collection and expenditures more
progressive (without sacrificing growth).
CGD0502 0527_Engl_6x9.indd 9
8/17/06 2:54:56 PM
10
Rescuing the World Bank
Bring New Discipline and
Greater Differentiation to Low-Income
Country Operations
Support for an expanded Bank role in low-income countries
is broad-based. At the same time there are widespread
doubts about its past effectiveness in these countries, many
of which have weak governments and limited absorptive
capacity, and failing to grow much in the past, acquired
unsustainable debt burdens. The new president should:
• Signal support for a much more differentiated
approach depending on each country’s governance,
in terms of the size and types of transfers, with
longer-term commitment periods for the bestperforming countries and much more flexibility in
reducing transfers (“exit”) when progress stalls, while
maintaining robust administrative spending to sustain
policy dialogue and engagement and technical
assistance in all countries independent of the size
of transfer programs.
• Urge the shareholders to formalize a third, fully grantbased window for countries with very low per capita
incomes, for example, below $500; most of these are
countries whose poor record of growth implies little
capacity to take on debt.
• Work with the International Monetary Fund on an
agreed role for the Bank in signaling the adequacy
of a country’s “development” approach and on a
facility to protect selected International Development
Association countries against external shocks.
Take Leadership on Ensuring Truly Independent
Evaluation of the Impact of Bank and Other
Aid-Supported Programs
Although the Bank has improved its level of transparency
through its research and the increasingly frank and
systematic work of its internal evaluation department,
neither satisfies the need for a credible, truly independent
assessment of the impact of development investments.
Echoing calls from the Meltzer Commission, the Overseas
Development Council Task Force on the Future of the
CGD0502 0527_Engl_6x9.indd 10
8/17/06 2:54:56 PM
11
A CGD Working Group Report
IMF, and the Gurría-Volcker Commission for independent
evaluation across donors, the Working Group recommends
that the president:
• Take leadership in working with the board to support
the creation of an independent evaluation entity
financed and governed by a consortium of public and
private donors and recipient countries, to complement
current internal audit and evaluation activities.
Obtain an Explicit Mandate, an Adequate
Grant Instrument, and a Special Governance
Structure for the Bank’s Work on Global
Public Goods
Over the years, the Bank has been drawn into the financing
and provision of a multitude of global programs ranging
from the environment to public health. The result is a
situation in which the Bank has a set of ad hoc global
programs without a clear mandate from its shareholders
and without the grant instrument needed for its more
effective engagement in provision and financing of
high-priority global public goods. The Working Group
recommends that the president:
• Call on the shareholders to develop a clear mandate
for the Bank’s role in the financing and provision of
global public goods.
• Initiate and maintain an ongoing dialogue with the
regional development banks, the United Nations,
and other relevant agencies to develop the proper
division of labor for respective work on global and
regional public goods.
• Call on shareholders to create a Global Public Goods
Trust Fund to finance the Bank’s work on global
public goods, based on agreed annual transfers from
the Bank’s net income and on contributions from
non-borrowers. Propose a governance structure
for the trust fund ensuring at least 40 percent
representation of middle-income and emerging
market economies (whose borrowing contributes
to net income) and 10–20 percent representation of
International Development Association countries.
CGD0502 0527_Engl_6x9.indd 11
8/17/06 2:54:56 PM
12
Rescuing the World Bank
Push the Bank’s Member Governments to Make
the Bank’s Governance More Representative and
Thus More Legitimate
The Bank’s own governance fails to adequately represent
the contribution and the interests of its borrowing
members. The lack of adequate representation is
undermining its legitimacy and puts its effectiveness at
risk. Yet there is no issue that has been as impervious
to change. The president should:
• Request that the governors of the Bank discuss
and formalize a mechanism for choosing the
Bank’s next president that is credible, rule-based,
and transparent.
• Support establishing two additional seats on
the board for African countries, pending a larger
consolidation to fewer restricted board seats.
• Ask the Bank governors to call for an independent
and public assessment of voting shares and board
representation, including assessment of the merits
of double-majority votes on selected issues and
taking into account discussion in the current quota
review at the International Monetary Fund of its
quota allocation.
• Ask the governors to commission a time-bound
independent review of board functions and
responsibilities, with an eye toward increasing
its overall effectiveness in holding Bank
management accountable.
CGD0502 0527_Engl_6x9.indd 12
8/17/06 2:54:56 PM
The Hardest Job
in the World
P
aul Wolfowitz assumes the presidency of the World
Bank at a key moment for the Bank and for the
development community. The Bank, as the world’s
Conflicting
premier institution for development, is to play a big part in
demands
the success of a revitalized global consensus—formalized
from mutiple
in 2000 by more than 150 heads of state at the United
quarters
Nations—to halve global poverty and to reach many other
make it
Millennium Development Goals by 2015. Why? Because
impossible to
of its financial strength and because of the breadth
and depth of its staff’s expertise on a wide range of keep all constituencies
development issues.
happy
But the Bank faces some real challenges in adapting
its internal governance structure and instruments—put
in place 60 years ago—to dramatic changes in the
global economy and in the relative power and needs
of its shareholders. The rise of China, the creation of a
European Union, the dramatic increase in private capital
flows to developing countries, the new risks of AIDS and
global terrorism—all are telling examples.
As a public institution, the Bank relies on the financial
and political support of its government members—
and, in the new global environment, of many other
constituencies and “stakeholders.” Yet conflicting
demands from multiple quarters make it impossible to
keep all constituencies happy.1 It is thus an easy target.
Those on the left accuse it of protecting privileged insider
financial and corporate interests—and perpetuating the
influence of the United States and other G-7 members
rather than the world’s poor people and their civil
society supporters. Those on the right accuse it of
misusing public resources in emerging markets where
private markets could operate better—and creating aid
dependency in the poorest countries where its loans
have contributed to unsustainable debt.2
The poorest countries, especially in Sub-Saharan
Africa, have called on the Bank to dramatically increase its
operations to help them meet the Millennium Development
13
Goals. Yet inside as well as outside its walls, there are
CGD0502 0527_Engl_6x9.indd 13
8/17/06 2:54:57 PM
14
Rescuing the World Bank
serious concerns about its effectiveness in the many
poor countries whose state systems are still weak or,
worse, corrupt. Meanwhile Bank operations are declining
in China, India, and the big middle-income economies,
as they borrow less and less. But the Bank’s mainstream
lending to these countries is what sustains its in-house
expertise and helps finance its administrative budget,
some of its poor country programs, and many of its
nonlending activities.
Ironically, because of its financial resources and its inThe biggest
house management and expertise, the Bank’s members
challenge
expect it to respond to multiple demands to do everything,
for the
from assessing post-conflict reconstruction needs in
Bank’s new
Kosovo and Iraq to developing a pilot program for trading
president
carbon emission rights across borders, to coordinating
will not be
closely with other donors and the United Nations on
managing
the Millennium Development Goals. Simultaneously, its
the Bank
management is accused of “mission creep.”3
but providing
No surprise, then, that the Bank is under pressure.
global
Its legitimacy, its credibility, its effectiveness, and its
leadership
fundamental mission are all in question—as is its future
in the fight
stream of support and income. In his decade as president,
against
James Wolfensohn managed several of those pressures
poverty
quite deftly. But without agreement of the Bank’s member
governments to fundamental changes in its governance
and the instruments at its disposal, he could not address
them all.
The Bank’s new leader needs to be ambitious. He faces
an unusual risk—for all the Bank’s strengths, merely
continuing with business-as-usual risks undermining its
future. He also faces an unusual opportunity—to provide
global leadership in advancing the global development
project. In the Bank’s self-effacing bureaucratic parlance,
global leadership is called “working with the shareholders”—
the Bank’s “shareholders” are the nations of the world.
This report defines a forward-looking agenda for the new
World Bank president to tackle over the next five years.
Its recommendations focus on five crucial tasks. They are
tasks for which the president—through a combination
of charm, cajoling, and horse-trading—must corral the
Bank’s recalcitrant collective of member governments,
including its single largest shareholder, the United States,
to take action—action to strengthen and secure the Bank’s
credibility, legitimacy and effectiveness for the twentyfirst century.
CGD0502 0527_Engl_6x9.indd 14
8/17/06 2:54:57 PM
15
A CGD Working Group Report
Guiding Principles
The five recommendations in this report are informed by
the following guiding principles.
The Bank’s mission and in-country priorities
The Bank’s mission is to reduce poverty in developing
countries. The most effective path to poverty reduction
is economic growth that is equitable enough to reach
poor people. Growth should also be sustainable (in the
environmental sense), and, to be sustained over time,
driven by private sector investment. On these points there
is no real disagreement.
But a statement of the Bank’s mission does not alone
provide guidance on its own operational priorities (where
“operations” refer not only to loans and grants, but also
to “dialogue” and advisory services). In today’s complex
donor system, the Bank need not and should not do
everything everywhere. But without a clear mandate
to set country priorities, history and habit suggest that
Bank staff will continue doing just that—limited only by
borrowers’ willingness and ability to borrow.4 Although
the Bank naturally provides advice and loans in a wide
variety of areas across countries, it needs to be clear
on its own priorities within individual countries—often in
the complicated context of other donor programs. That
means ending the confusion between what is good for
development in general (such as girls’ education) and what
is good for equitable and sustainable growth in a particular
country at a particular time (where and when it might be
rural roads that have the highest marginal benefit—even
for encouraging girls’ education). And it means setting
priorities for what the Bank itself should do in each country;
even when the Bank is the single agency with the broadest
overall knowledge of a country’s development needs (which
is often but not always the case), it need not be the largest
financier of development investments.
Compounding the lack of clear operational priorities
in countries is a new round of uncertainty about the
ingredients of growth that can reduce poverty. Under
pressure from critics, Bank staff in the 1990s interpreted
the “poverty” mission as a mandate to lend directly for
poverty reduction. Combined with the pressures of
safeguards against environmental abuse in infrastructure
projects and the growing concerns about corruption in
CGD0502 0527_Engl_6x9.indd 15
8/17/06 2:54:57 PM
16
Rescuing the World Bank
the procurement process for large projects, the poverty
emphasis led to a shift of lending toward the social sectors.
In the last year or so, in marked contrast, there is renewed
talk of infrastructure as a priority, as a quicker path to
“growth” (and through growth to poverty reduction) than
social spending, and as less vulnerable to the bottlenecks
that management and human resources limits put on
rapid expansion of health and education systems.5
The development community has learned that no single
recipe or set of priorities to achieve poverty-reducing
growth can be applied across countries. On the one hand
there is broad consensus on the prerequisites of sustained
growth ranging from the importance of human capital,
in particular health and education, of macroeconomic
stability, and of institutions and governance. But
chastened by heterodox China’s spectacular growth,
reformist Latin America’s dangerous vulnerability to
external volatility, and Sub-Saharan Africa’s embarrassing
accumulation of unsustainable debt, the development
community and the Bank are less confident about how
precisely to operationalize broad, widely agreed upon
goals, in particular settings. Put another way, there is
no longer anything that could be called a “Washington
Consensus,” nor across all Bank borrowers, any simple
choice of encouraging more infrastructure versus more
social investment.6 There is, at best, a growing consensus
that sound institutions—political and economic—matter
and that institutions have to be invented locally, tailored
to local political and social realities. That puts a premium
on respect for and partnership with local efforts by Bank
staff, and on the need for country-specific knowledge
and expertise.7
Equitable growth and political savvy
The new emphasis on local institution building and local
ownership raises an additional challenge for the Bank.
Ownership and the loss of faith in any universal policy
package imply that the Bank should become less of a
“nanny” Bank, preoccupied with the detailed conditionality
and structural reform demands that dominated lending in
the late 1980s and much of the 1990s. It should instead
concentrate more on supporting healthy local economic
and political institutions. But local political ownership
in developing countries (indeed in all countries) is not
necessarily conducive to equitable or pro-poor growth.
CGD0502 0527_Engl_6x9.indd 16
8/17/06 2:54:57 PM
17
A CGD Working Group Report
The Bank’s engagement and financing in borrower
countries is supposed to leverage policies that are equityenhancing and public spending that is pro-poor. The result
is tension between a nanny Bank exercising leverage and
a politically naïve Bank overdoing country ownership.
The Bank’s future effectiveness depends on managing
that tension well. That implies much more attention to
identifying and quantifying corruption risks, interest
group pressures, and other local political constraints
(and opportunities). It means helping to lead multilateral
The Bank
efforts to combat bribery (such as the Extractive
should
Industries Transparency Initiative). And it means setting
concentrate
operational priorities that take those risks into account.8
more on
It implies an approach that is both politically sensitive
supporting
and politically savvy.
It also means increasing the resources for data healthy local
economic
collection, measurement, and analysis of corruption,
and
political
transparency, the rule of law, the business environment,
institutions
and so on across countries. All this is necessary for the
Bank to be a global brain trust addressing the difficult
politics (and economics) of growth that is pro-poor.
The Bank as development’s brain trust
External resources are mainly fungible, and the Bank
need not and should not be the primary source of
development finance (either because other donors
provide major resources in poor countries or because
government revenue and private capital provide the bulk
of resources in middle-income and fast-growing emerging
markets). The implication is the need to distinguish clearly
between the Bank’s role in transferring financial resources
and its particular comparative advantage: its singularly
overarching overview of global opportunities, institutions,
and constraints, and of borrowers’ institutional and financial
capacity. That overview is grounded in broad-ranging
and deep staff knowledge and experience on technical,
sectoral, and economic issues. No other institution has
the same strength in the generation and diffusion of
knowledge about the practice of development. (Indeed, as
a “knowledge bank,” creating and sharing across countries
development experience and expertise, the World Bank
itself constitutes a global public good—an institution that
no one country today would have sufficient incentive to
create or fund, yet from which all potentially benefit.)
CGD0502 0527_Engl_6x9.indd 17
8/17/06 2:54:57 PM
18
Rescuing the World Bank
Strengthening the brain
trust requires
the Bank
to retain its
financing role
The knowledge bank will be handicapped if not
supplemented by two additional efforts. First, it must
do more to create capacity for knowledge generation in
borrowing countries. The Bank cannot be a substitute for
independent policy thinking in borrowing countries—this
tendency has fueled the perception of a parochial and
arrogant institution. Second, it must make much greater
efforts at disseminating knowledge—as a knowledge
clearinghouse—for example, by such apparently basic
steps as making its website more multilingual.9
The Working Group acknowledged that strengthening
the brain trust requires the Bank to retain its financing
role. The lending process often triggers and supports
the policy dialogue and advice to countries, reinforcing
the Bank’s capacity as a brain trust. Advice not linked
to finance too often ends up on ministry bookshelves.
The income from lending also augments the resources
to support the institution’s advisory function.
The Bank’s governance: toward greater
legitimacy and effectiveness
The Working Group’s recommendations look toward a
transformation of the Bank from a development agency—
in which some members are financial contributors and
others are beneficiaries—to something closer in spirit to
that of a global “club.” In a global club today’s developing
country beneficiaries, not only its rich country benefactors,
would have a keen sense of ownership and financial
responsibility. Such a transformation would recognize that
the Bank cannot be effective and relevant in addressing
major global economic problems if countries such as Brazil,
China, India, South Africa and Turkey are not full members
with corresponding rights and responsibilities.
The Working Group also noted that improving
the Bank’s governance was only part of the larger
challenge of building a more legitimate and effective
overall system of global governance, encompassing the
Bretton Woods institutions, the regional banks, the World
Trade Organization, and the United Nations. Improved
World Bank governance should thus be seen as part of
broader and deeper reforms of the existing international
architecture. The World Bank and the International
Monetary Fund (IMF), in particular, must work to build
a more constructive and effective partnership, not only
with each other but also with the United Nations.
CGD0502 0527_Engl_6x9.indd 18
8/17/06 2:54:58 PM
19
A CGD Working Group Report
Five Crucial Tasks
Drawing on these guiding principles, the Working Group
identified five crucial tasks for the new president. These
tasks are not meant to be comprehensive. They are tasks
where the president’s leadership is needed to guide and
shape decisions by the Bank’s member governments
and where the absence of leadership risks undermining
the Bank’s contributions going forward. The five crucial
tasks are:
A transfor-
• Revitalize the Bank’s role in China, India, and middle- mation of the
income countries.
Bank from a
• Bring new discipline and greater differentiation to development
low-income country operations.
agency to
• Take leadership on ensuring truly independent
something
evaluation of the impact of Bank and other aidcloser in
supported programs.
spirit to that
• Obtain an explicit mandate, an adequate grant
of a global
instrument, and a special governance arrangement
“club”
for the Bank’s work on global public goods.
• Push the Bank’s member governments to make the
Bank’s governance more representative and thus
more legitimate.
Revitalize the World Bank’s Role in China,
India, and the Middle-Income Countries
The Bank’s role in middle-income countries and in such
low-income but fast-growing emerging markets as China
and India can no longer be taken for granted. The majority
report of the International Financial Institution Advisory
Commission (mandated by the U.S. Congress in 2000
and commonly referred to as the “Meltzer Commission”
for its chairman, Allan Meltzer) recommended that the
Bank stop lending to emerging market economies and
middle-income countries with ready access to private
capital markets.10
Many of the Bank’s fast-growing and middle-income
borrowers seem to share this view. The long-term trend of
their borrowing from the Bank is clearly down, especially
from the exceptionally large outflows of the Bank in the
late 1990s during the financial crises that hit East Asia,
then Russia, and then Brazil and Argentina. Bank staff and
country officials generally take the view that the decline
reflects reduced demand from borrowers, not reduced
willingness to lend.11
CGD0502 0527_Engl_6x9.indd 19
8/17/06 2:54:58 PM
20
Rescuing the World Bank
The World
Bank should
continue to
be active
in middleincome
countries
and in such
emerging
markets as
China and
India
CGD0502 0527_Engl_6x9.indd 20
Why lend to countries with access to
private capital?
The Working Group concluded, in contrast, that the World
Bank should continue to be active in middle-income
countries and in such emerging markets as China and
India.12 There are at least three reasons: many of these
countries’ limited access to private capital markets,
the legitimate interest of the Bank’s rich members in
encouraging pro-poor, equitable growth policies in middleincome and emerging market economies, and the logic
and evident success of past “bundling” of policy advice
with loans.
Easing limited access. Even for the large countries
that are deemed attractive to investors, private capital
markets (internal and external) are still volatile and procyclical. For the poorer middle-income countries (such
as Guatemala, Kazakhstan, and Paraguay) as well as
those where internal conflicts persist (the Philippines and
Sri Lanka) and domestic debt is high (Brazil and Turkey),
access to external capital is still largely limited to shortermaturity loans. For these and other richer economies in
this category access to internal capital is often costly
due to relatively weak and shallow banking systems,
small, illiquid local capital markets, and the risk that too
much sovereign borrowing in thin domestic market will
make banking systems more vulnerable.13 In almost all
of the Bank’s middle-income borrowers, only time and
performance—much more than a decade of steady, sound
economic policies—and the visible resilience of economic
and political institutions will induce domestic and foreign
creditors and investors to accept lower returns for their
capital in return for lower country risk.
Experience shows, moreover, that the cost and
availability of funds in international markets can change
abruptly, sometimes for reasons beyond the control of any
country. In the process, economic growth, development
strategies, and antipoverty programs may suffer setbacks.
When global turmoil partially or completely closes market
access, multilateral lending can assist in sustaining
adequate public spending on education and health, in
strengthening regulatory and supervisory capacity, and
in developing social safety nets—as in Mexico in 1995
and the Republic of Korea in 1998. Since crises tend
to hurt the poor the most through lost employment and
8/17/06 2:54:58 PM
21
A CGD Working Group Report
income and interrupted education for children, assisting
countries in coping with crises helps alleviate poverty and
promote development. When the Bank maintains and even
increases lending during periods of stress, it signals support
for responsible development policies, and with relatively
modest amounts helps rebuild market confidence.
In the meantime, longer-term and cheaper loans
from the World Bank can encourage public investments
with high social and economic returns that do not yield
commercial returns to private agents (such as investments
in education, health, rural infrastructure, bank regulation,
and judicial reform) and that otherwise might not find
a place in national budgets. These are investments
that, by supporting equitable growth in open market
systems, create an environment that crowds in productive
private investment.14
Promoting equitable growth. Even putting aside volatile,
crisis-prone access to capital, advanced economies have
an interest in reducing poverty in developing countries
and in investing in human resources. The Working Group
noted that more than two-thirds of the world’s poor lives in
middle-income and emerging market countries. China and
India alone account for 45 percent of the total. Pro-poor
and human development instruments yield high returns
but only in the medium term, and countries with weak tax
systems cannot easily translate the economic returns for
a road into the tax revenue to repay short-term loans.15
The social and economic decisions of middle-income
countries affect the health and well-being of their own
peoples, undermining or advancing such global goals
as poverty reduction.
Moreover, the United States and other nonborrowing
members have a substantial security stake in the
institutional resilience of middle-income and emerging
market economies. Their financial stability contributes
to global financial stability. And their decisions—on
commodity and energy use, international capital market
borrowings, reducing corruption, and so on—affect the
once-insulated residents of rich countries. For that reason
nonborrowing members also have a legitimate interest
in encouraging middle-income and emerging market
economies to invest in programs that generate global
and country-specific benefits.
CGD0502 0527_Engl_6x9.indd 21
8/17/06 2:54:58 PM
22
Rescuing the World Bank
Bundling policy advice with loans. Lending operations are
a vehicle for supporting and rewarding policy reforms and
development results. And there are good reasons to doubt
that unbundling the financing from the “dialogue” about
policy and results would always be effective (though it
may make sense for some countries). The Bank does
and should continue to charge for advisory services,
proving the worth of its stock of expertise. However,
political and social constraints in emerging markets, as
well as technical complications, make it difficult to design
Lending
and implement many reforms—for example of health,
operations
banking systems, bankruptcy law, and pension and
are a
unemployment programs. Officials from such countries as
vehicle for
Brazil, Hungary, the Republic of Korea, Mexico, Thailand,
supporting
and Turkey repeatedly cite the services bundled with Bank
and reward- financing as a key reason for seeking Bank loans. They
ing policy
cite the leverage that the potential financing provides them
reforms and within their own political settings as helpful in persuading
development and encouraging progress on their reform agenda. They
results...
value not only the dialogue on tough internal policy and
budget choices that the lending process catalyzes but
also the detailed, project, sectoral, and economic analysis
by Bank staff.16
Services bundled with lending also support objectives
of the global community: human development, protection
of the environment, financial accountability, and standards
of public procurement that curtail corruption and promote
competition. For example when Bank financing supports
general government expenditure, the accompanying
dialogue and advice promote better debt management
and responsible budget management. Put another
way, lending is the vehicle for the Bank, by supporting
reformers within government, to influence governance
issues (accountability of government, and greater
representation of all citizens in economic decisionmaking)
and contribute to strengthening democratic institutions
in emerging market economies.
Then why are these countries borrowing less
and less?
What lies behind the decline in the demand for Bank
loans? Despite the below-market interest rates and long
maturities of Bank loans, the trend in the last 15 years
has been for middle-income borrowers to reduce their
new borrowing from the Bank. In some cases, countries
CGD0502 0527_Engl_6x9.indd 22
8/17/06 2:54:58 PM
23
A CGD Working Group Report
have prepaid loans whenever the cost of borrowing on the
private market has been low. For fiscal years 1990–97,
International Bank for Reconstruction and Development
(IBRD) lending, measured by gross disbursements, was
in the range of $15–18 billion (figure 1). There was a brief
spike in response to the Asian financial crisis, but in fiscal
2004 lending dropped to the $10 billion mark. As a result,
for many middle-income borrowers, net transfers from
the Bank are now negative.
To some extent, this trend is healthy. Some one-time
...and by
borrowers (Hungary, the Republic of Korea, Malaysia,
supporting
Singapore, and Thailand) have graduated from Bank
reformers
borrowing—though some returned when hit by the
within
global financial problems of the late 1990s. To some
government,
extent, the emerging market economies’ vulnerability to
global financial volatility has led to a tougher standard strengthening
democratic
on acceptable external debt-to-GDP ratios for such
institutions
countries—as low as 40 percent—which has reduced
in
emerging
demand for external borrowing in general. But developing
market
countries should generally be net importers of capital not
economies
exporters. (The illogic of negative net transfers from the
Bank is repeated and dramatized in the illogic of some
emerging markets accumulating large dollar reserves.)
It is also true that the Bank appears to be succumbing
to the broader problems rather than compensating for
them, and that officials of the middle-income countries
and emerging market economies have additional reasons,
more Bank-specific, for their declining demand for Bank
loans. One is the limitations of the Bank’s longstanding
main product: the sovereign guaranteed loan, compared
Figure 1 IBRD Lending, Fiscal 1995–2004
$ billions
25
Gross disbursements
Commitments
Adjustment lending
20
15
10
5
0
1995
1996
CGD0502 0527_Engl_6x9.indd 23
1997
1998
1999
2000
2001
2002 2003 2004
8/17/06 2:54:59 PM
24
Rescuing the World Bank
with the range of innovative financial products available
in the private market. A second is the high “hassle” or
transaction cost (not the financial cost) of borrowing
from the Bank, compared with the small and declining
difference in cost of borrowing from private markets in
recent years.
Within the World Bank Group, the International Finance
Corporation (IFC), the Multilateral Investment Guarantee
Agency (MIGA), and the Foreign Investment Advisory
Service (FIAS) do supplement the IBRD’s sovereignguaranteed loans with lending, equity investments and
advisory services, primarily to private sector agents. But
it is still an uphill battle to see any common strategic
direction built into country programs of these various arms
of the Bank Group. And the fact remains that the financial
capacity and the balance sheet of the IBRD are much
larger than those of the IFC, and the membership and the
financial clout of MIGA and FIAS remain limited. As a result,
the World Bank Group continues to lag behind in the range
of its products, and the IBRD has limited means to make
meaningful its allegiance to private sector growth, since
its main instrument requires a sovereign guarantee.
For example, the IBRD does offer partial risk guarantees
as well as loans.17 But because the IBRD’s financial policies
require that these guarantees be priced and provisioned in
virtually the same manner as loans, the guarantee is not as
attractive to the countries as a loan, and demand for it has
been close to zero. In addition, the Bank’s guarantees also
require a sovereign counterguarantee. But the requirement
for a counterguarantee violates the reasonable requirement
of responsible central governments to avoid backing up
subsovereign borrowing.
Similarly, the IBRD cannot make loans to municipal
and other subsovereign governments without a formal
guarantee of the central government—which as with
guarantees many governments now eschew, since such
guarantees undermine the accountability of nonsovereign
political entities and thus the healthy development of
disciplined local government. (The regional development
banks do not have the requirement for the sovereign
guarantee built into their charters, and in the last
decade they have begun to offer a broader range of
products, including those for subsovereign and private
borrowers. As a result, the World Bank Group may be
CGD0502 0527_Engl_6x9.indd 24
8/17/06 2:54:59 PM
25
A CGD Working Group Report
losing its longstanding position of leadership in analysis
and innovation.)
In short, credit products come in forms other than
loans, and the Bank could add more value for some
of its borrowers by going up the credit-product value
chain (just as commercial players have done with their—
more limited—appetite for emerging market risk). New
credit products could also have more of an insurance
element to them, insuring against such market-related
risks as movements in interest rates, foreign exchange
The Bank
rates, and commodity prices, which the private sector
could
add
structures and distributes, but for which their credit
more
value
appetite is limited.18
for some of
The “hassle” problem that discourages borrowing
its
borrowers
includes the long lapse of time between a government’s
by
going up
initiating a loan request and getting the loan approved, the
the
creditonerous administrative burden of preparing, negotiating,
product
value
and implementing Bank-financed programs and projects,
chain
and the administrative and financial costs of dealing
with the growing demands of the Bank—often pushed
by well-meaning civil society groups in the advanced
economies—that borrowers meet high environmental and
other standards in the design and implementation of Bankfinanced projects. Whether these standards are “too high”
is a matter of controversy. That they raise the perceived
if not actual costs of projects and can slow down their
approval and implementation is undoubted.19
Without the very large and fast-disbursing “adjustment”
loans to countries during the Asian financial crisis, to
Brazil in 1999, and to Argentina in 2000–01 to supplement
IMF balance of payments support, the Bank’s net income
from its “bread and butter” loans would be even lower.
Indeed, one reason countries prefer adjustment loans is
that they dramatically reduce up-front “hassle” costs, and
they eliminate the resource and reporting burden of the
“counterpart” funds that countries are usually required to
provide from their own budgets for conventional projects.
With the requirement for such counterpart funds, the Bank
also compromises its effectiveness as a countercyclical
lender, which would otherwise help countries minimize
the social costs of economic downturns.
Why does it matter?
The demand for borrowing from this set of countries may
on current trends remain low and fall further—despite
CGD0502 0527_Engl_6x9.indd 25
8/17/06 2:55:00 PM
26
Rescuing the World Bank
the benefits to the borrowers and despite the legitimate
security and development interests of nonborrowers in
the Bank’s continuing engagement with those countries.
Over the next decade, a rapid exit of more creditworthy
borrowers poses three additional risks to the Bank: a severe
adverse selection problem in the Bank’s portfolio, reduced
net income, and lost opportunities for the Bank to transfer
experience from middle-income countries to low-income
countries as they develop.
The Bank’s own cost of borrowing might rise slightly
Major
if its creditors saw greater portfolio risks, reducing the
changes are financial benefits to the very countries still in most need
needed in the of Bank financing. Equally problematic for all the Bank’s
operations
members would be a reduction in its net income—its
of the Bank
income from the spread between its cost of borrowing
if it is to be
and the interest earned on its loans. Low demand from
effective and the Bank’s IBRD borrowers risks undermining not only
relevant in
its potential positive role in middle-income countries but
middle-income also the financial strength on which its other roles—in
countries
low-income countries, in transferring cross-country
experience, in providing advice, and in supporting the
provision of global public goods—at least partly depend
(figure 2). It also risks reducing the ongoing internal
learning and knowledge-generating role of Bank staff,
who learn in an active lending program.
We conclude that major changes are needed in the
operations of the Bank if it is to be effective and relevant
in this group of countries. Our recommendations to the
new president for China, India, and the middle-income
countries are as follows:
1. Ask the shareholders to begin a systematic and careful
review of whether the charter requirement that IBRD
loans be guaranteed by a sovereign borrower has
stifled the Bank Group’s ability to respond to the
changing demands of its key borrowers.
Could it be that the separate balance sheets of the
IFC and the IBRD, for example, have discouraged
the development of new products to catalyze private
sector investment in middle-income and emerging
market economies? Does the IBRD charter make
it too difficult for the Bank to lend to municipal
and other subnational and other subsovereign
government entities?
CGD0502 0527_Engl_6x9.indd 26
8/17/06 2:55:00 PM
27
A CGD Working Group Report
Whether or not the outcome of any such review
would lead to structural changes, the Working Group
believes it would open the door to new thinking
about the medium-term instruments that the Bank
and other multilateral development banks need to
be responsive to the key problems and changes in
the global economy.
2. Find ways, within current constraints, to sharply expand
the range of financial products and instruments now
available to borrowers. It is widely acknowledged
Find ways
that the Bank has been extremely innovative when it
to sharply
comes to its own borrowings and investments. But it
expand the
has been anything but innovative in its own product
range of
offerings, which remain almost entirely concentrated
financial
on the single-priced sovereign guaranteed loan. products and
Examples of possible new products and related new instruments
approaches include:
now available
• Risk management products and instruments to
hedge against commodity risk. (In emerging market
economies the private sector has little appetite for
providing these services, and the Bank could step in to
fill a clear void.) Risk-sharing loan contracts could tie
the rate of interest on sovereign loans to commodity
export prices, especially for countries heavily
dependent on primary commodity exports.
• Leveraging the Bank’s financial strength and
shedding the undue conservativeness that have
to borrowers
Figure 2 IBRD Net Income: Sources and Uses,
Fiscal 2004
Uses
Sources
Surplus
$450 million
Borrowing
Interest margin
on loans
Securities and
other financial
instruments
Interest on
investments
Paid-in capital
Contributions
from equity
Reserves
CGD0502 0527_Engl_6x9.indd 27
Reserves
$680 million
Heavily Indebted
Poor Country Initiative
$240 million
International Development Association
$300 million
Administrative
expenses
$1,100 million
Other
$50 million
Global public
goods
$179 million
8/17/06 2:55:01 PM
28
Rescuing the World Bank
Create a new
loan product
that would
visibly reduce
hassle costs
for selected
borrowers
made Bank guarantees no more attractive than
loans, while tightening the distinction between
guaranteeing political risk, which the Bank should
do, and commercial risk, which it should avoid.
• Borrowing in local capital markets to help
strengthen these markets and lending in the
local currency (ideally long-dated, fixed-rate,
and indexed to local price levels so that the debt
cannot be inflated away) to help borrowers avoid
the currency risk that borrowing from the Bank
usually entails.
• Developing other products to help borrowers
reduce their currency risks. The Bank could, for
example, borrow in a synthetic unit whose value
was determined by a basket of inflation-indexed
emerging market currencies, and sell bonds
denominated in this unit to international investors.
The Bank would cover itself against exchange risk
by on-lending the borrowed money to countries in
their own currencies (on an indexed basis, in the
proportions that make up the basket).20
• Working with the IMF to explore still other
possibilities, for example, on how Bank lending
could contribute to refinancing the sovereign
debt of overindebted middle-income countries
(at marginally more than the Bank’s borrowing rate
and ideally in a country’s own currency), in return
for continuing, monitored progress on disciplined
macroeconomic priorities.21
3. Create a new loan product that would visibly reduce
hassle costs for selected borrowers.
The effect of the profusion of project safeguards
and program conditionalities on quality may
be driving away some borrowers, particularly
from interest in large infrastructure projects. For
borrowers with reasonably good performance in
economic management and an adequate record
and regulatory effort in procurement, environmental
protection, and human rights, the Bank should move
to a more arm’s length relationship. The Working
Group recommends that the Bank develop a new
instrument that would greatly reduce the hassle cost
for creditworthy countries that are vulnerable when,
CGD0502 0527_Engl_6x9.indd 28
8/17/06 2:55:01 PM
29
A CGD Working Group Report
for example, new investments require resettling of
people in new locations.
An existing facility—the “deferred drawdown
option”—is a start in this direction, but has not been
attractive to borrowers because it is not clear that it
would be sufficiently automatic. We recommend that
the Bank develop few and well defined standards of
eligibility, developed in consensus with all members,
and that the list of countries with eligibility for onestop access be updated periodically. Terms of
Add a
eligibility could be revisited and redefined every
degree
of
three years or so. The reduction in the “hassle
differential
factor” would not only increase the demand for
pricing
Bank loans—even assuming a higher borrowing
among
IBRD
rate for eligible countries—but would also reduce
borrowers,
the Bank’s administrative costs. The latter savings,
tied strictly
as we argue later, would yield substantially greater
to
per capita
social returns if deployed in the financing of global
income
public goods.
Not all middle-income countries will meet the
eligibility requirements for such hassle-free lending,
and we are certainly not suggesting that only
countries that are high-performing be eligible for
any loans (as with the U.S. Millennium Challenge
Account). The Bank should continue to take risks
in middle-income countries where resources have a
reasonable probability of being used effectively and
where conventional monitoring and conditionality
can increase that probability.
4. Add a degree of differential pricing among IBRD
borrowers, tied strictly to per capita income (not to
credit rating), recognizing that the implicit benefit to
less creditworthy borrowers is already larger than to
more creditworthy ones.
A marginally higher rate for richer countries
with better credit ratings would encourage less
borrowing for the right reason—ushering in de facto
“graduation” without any recourse to arbitrary rulebased loss of access.22 It would also, like the facility
proposed earlier, create incentives within the Bank
to reduce hassle costs for the somewhat better-off
middle-income countries—some of which are now
fully capable of preparing and managing large loans
CGD0502 0527_Engl_6x9.indd 29
8/17/06 2:55:01 PM
30
Rescuing the World Bank
for electricity distribution, agricultural research, and
health systems.
5. Explore other pricing or product innovations that would
create incentives for borrowers to make their own public
revenue collection and expenditures more progressive
(without sacrificing growth)—and that would encourage
investments with a high payoff for global public goods.23
Starting from a country borrowing rate based on per
capita income, loan charges could be reduced for large
ramp-ups in expenditures on financially high-risk but
clearly pro-poor sectors, such as basic education and
health, rural roads, training recipient country nationals,
and other long-term capacity building.
In most middle-income and emerging market
economies, there is no tradeoff between the
government’s fiscal behavior being more equitable
and at the same time more efficient. Indeed, fairer and
more equitable revenue and expenditure patterns
would be more efficient—for example, because public
spending on health and education of reasonable
quality increases worker productivity, and because
reduced tax evasion and lower trade and payroll
taxes are both pro-poor and growth enhancing.24
Because resources are fungible, clear rules on the
increment to the proportion of government budgets
to be eligible for this kind of incentive would need
to be developed. This approach would also require
clear rules of country eligibility.
Bring New Discipline and Greater
Differentiation to Low-Income
Country Operations
Support for a strong and even expanded Bank role in
low-income countries is broad-based. This is especially
true for Africa, particularly in the context of the 2005 U.N.
Millennium Review Summit (which will evaluate progress
toward the Millennium Development Goals) and the United
Kingdom’s call to address Africa’s problems at the 2005
G-8 Summit in July. Reflecting that support, rich-country
contributions to the Bank’s International Development
Association (IDA) window increased from $13 billion in
the 13th replenishment to $18 billion in the 14th, and
bilateral foreign aid commitments from Europe and the
United States have surged in the last several years.
CGD0502 0527_Engl_6x9.indd 30
8/17/06 2:55:01 PM
31
A CGD Working Group Report
Discipline and differentiation
Broad support for the Bank’s engagement in helping lowincome countries achieve the Millennium Development
Goals should not obscure concerns about the Bank’s
and other donors’ effectiveness in those countries.
These concerns range from the difficulty of avoiding
imposing ideas and recipes (as reflected in the view that
Poverty Reduction Strategy Papers still reflect countries’
expectation of what the Bank wants more than their own
priorities) to the difficulty the Bank and other donors have
in “exiting,” in reducing their transfers when countries are
not using external help well.25
Donors are making substantial efforts to increase
their coordination and harmonize their approaches in
low-income countries, many of which receive aid from
dozens of bilateral and multilateral agencies as well as
international nongovernmental organizations (NGOs). The
Bank, often the most influential among many donors,
needs to set the tone—creating space for countries
to manage their own priorities wherever that makes
sense, cooperating with others in helping countries set
clear priorities, and ensuring in its own operations more
discipline and differentiation in the amount and nature
of support it provides, depending on recipient countries’
capacity, governance, and economic management.
This will require substantive changes in the way the Bank
does business. The Bank, with its vast array of expertise
on a wide set of issues, coupled with its decentralized
structure, has tended to encourage strategies and
programs on a wide set of initiatives, with no sense of
which are the most important.26
But most governments in low-income countries simply
do not have the capacity to tackle a very wide agenda.
Governments with very scarce time, money, and skilled
staff need to set priorities, which in practice usually means
deciding which issues will not be dealt with right away.
The Bank, as one of many partners, including the relevant
regional development bank, U.N. agencies, bilateral
donors, and international and local nongovernmental
organizations—should have a broad strategic role advising
a country on its priorities—combined with what should
often be a narrow focus for its own lending.
The Bank should also take more of a lead in helping
donors discriminate across low-income countries in the
amount and nature of their transfers. The discussion in
CGD0502 0527_Engl_6x9.indd 31
8/17/06 2:55:01 PM
32
Rescuing the World Bank
The Bank
should take
more of a
lead in
helping
donors
discriminate
across
low-income
countries
recent years about “country selectivity” has led to the
idea of providing large sums of money to well-governed
countries that can use it well and less to poorly governed
countries (the U.S. Millennium Challenge Account’s
approach). But well-governed countries should not
only receive more money, they should receive it in more
attractive ways that give them more substantive input,
responsibility, and certainty about future funding. The Bank
has moved tentatively in this direction by funding SectorWide Approaches and introducing Poverty Reduction
Support Credits in certain countries. But these different
approaches should become more formalized.
In less well-governed countries, the Bank should be
much more modest—limiting its lending and limiting its
expectations. It should not reduce its engagement, its
budget for policy dialogue and technical assistance, or
its willingness to take certain risks. On the contrary, the
administrative budget for poorly performing countries
should be explicitly untied from the program of lending or
grants. But the Bank should be more prepared to suspend
financing where that makes sense and to design programs
that build in such suspensions when progress stalls.
Specifically, the Bank should have three distinct
strategies for low-income countries, depending primarily
on the quality of the recipient’s governance.27
1. Low-income better governed countries. The Bank
should provide large amounts of financing to these
countries, delivered mostly in the form of budget
support or program aid. Along with other donors, it
should focus less on micromanaging activities and
more on measuring and achieving broad results. The
Bank should commit funding for five years or more
in these countries, subject to the strict requirement
that recipients show continued good governance and
achieve reasonable results.
2. Low-income countries with average governance.
These countries should receive less funding than the
better-governed countries. The Bank should be more
involved in setting priorities and ensuring broad-based
participation and technical rigor. Strengthening public
financial management is usually a very high priority
in these countries, and to strengthen reforms in this
area some budget support may be appropriate. Most
funding should, however, be for well-designed projects
CGD0502 0527_Engl_6x9.indd 32
8/17/06 2:55:02 PM
33
A CGD Working Group Report
or sector support consistent with the country’s overall
development strategy, focusing on key activities where
achieving results seems most likely, well integrated
with country budgeting and financial management
arrangements. The length of financial commitments
should be shorter than for well-governed countries,
perhaps three to five years, contingent on progress
and results. Performance should be monitored
carefully in these countries, with clearly delineated
performance standards. Strong performance and
improved governance should lead to increased
financial support, a shift to budget and sector support,
and longer commitments—while weak results should
lead to less aid.
3. Low-income poorly governed countries. These
countries, broadly consistent with the Bank’s “LowIncome Countries Under Stress” must be dealt with
carefully case-by-case, because circumstances
on the ground can vary widely and change quickly.
Some are failed states, others are failing, still others
are weak or fragile. Some donors (but probably not
the Bank) should direct significant amounts of aid to
civil society groups and NGOs.28 The Bank should
continue with substantial engagement and carefully
targeted technical assistance, but should not generally
be providing financing to government. It should not
get into retail-style grantmaking to civil society groups
because its comparative advantage is in working
directly with governments.
In less wellgoverned
countries,
the Bank
should be
prepared
to suspend
financing
and to design
programs
that build in
suspensions
when
progress
stalls
Grant financing
Differentiation on the amount of resource transfers should
be based primarily on country governance. Another kind
of differentiation—for the type of transfer (grant or loan)—
should be based on countries’ per capita income. Since
President Bush proposed that 50 percent of IDA funds
be used as grants in July 2001, there has been a strong
debate about the extent to which the Bank should provide
grants rather than loans to low-income countries. The
rationale for IDA shifting to greater use of grants is the
past accumulation of unsustainable official debt by many
low-income borrowers, including debt to the World Bank.
Much “new lending” prior to debt reduction was simply
helping countries repay former loans.
CGD0502 0527_Engl_6x9.indd 33
8/17/06 2:55:02 PM
34
Rescuing the World Bank
Negotiations between the United States and the
Europeans (who were concerned about the effects of
reduced future reflows for IDA’s finances) led to an initial
fuzzy compromise during the IDA-13 replenishment in
2002. The Bank’s Board decided that 18–21 percent of
IDA funds would be grants for a smorgasbord of purposes:
post-conflict reconstruction, natural disasters, HIV/AIDS,
education, health, water, and sanitation. This led to a
less-than-satisfactory outcome in which countries would
receive grants for some activities and loans for others.
The Bank’s
Recognizing these problems, the Board amended the
members
guidelines in March 2005 to make debt sustainability the
should agree basis for the allocation of grants. While an improvement,
to formalize
the arrangement is less than ideal. Using debt sustainability
a third, fully
as the basis for grants introduces moral hazard issues
grant-based (countries that have taken on more debt in the past will now
window for
receive grants, while those that have not must continue
countries
to borrow). It also creates administrative problems (doing
with very low country-by-country assessments of what portion of grant
per capita
financing each country should receive).
incomes,
Instead, grant allocations should be based primarily
for example, on income levels following the same principles that now
below $500
guide the allocation between IBRD and IDA loans. The
Bank’s members should agree to formalize a third, fully
grant-based window for countries with very low per
capita incomes, for example, below $500, an average
income just over the $1 day poverty line.29 The logic is
straightforward: loans make sense when the recipient’s
economy can grow fast enough to generate the resources
to repay the loans. But most countries with incomes
below $500 have never achieved sustained economic
growth—not for hundreds of years. Until they achieve
such growth, grants make far more sense than loans.
Moreover, the very poorest countries are least able to
cushion themselves against shocks, making it more
difficult to repay loans, even following good investments.
Given very scarce resources, any funds generated by
strong investments should be re-invested locally, not
repaid to the Bank.
Extending the IDA horizon for recipient
countries, while encouraging “exit”
when appropriate
Donors agree that for poor countries to meet the Millennium
Development Goals requires an increase not only in the
CGD0502 0527_Engl_6x9.indd 34
8/17/06 2:55:02 PM
35
A CGD Working Group Report
amount of aid but also in its predictability and horizon
over a longer period. Responsible finance ministers in the
poorest countries naturally hesitate to hire new teachers
and build new schools where the prospect of financing
their ongoing costs from the country’s own revenues is
limited, while external funds are volatile and uncertain
over the medium term. The Bank through its IDA window
should be more able and willing to make longer-term
commitments to the best-performing countries—as
long as 10 years—contingent on continuing progress
The Bank
against clearly defined benchmarks.30 The time horizon
should
be
for development in IDA countries is, after all, still 40–50
more
able
years. Consider Mozambique, with per capita income of
and willing
about $210. With very robust annual per capita income
to make
growth of 5 percent, it would take almost 30 years for
longer-term
Mozambique to reach the IDA operational cutoff of $865
commitments
per capita.
The Bank could extend the predictable horizon to the bestperforming
of its commitments. But it often errs in the opposite
countries
direction—prolonging commitments and programs
when countries are not meeting agreed benchmarks of
progress or are backsliding on human rights, on friendly
business environments, on expenditure management, or
on other measures of governance. Part of the problem
is the periodic pressure on Bank management, as a
result of the IDA three-year replenishment cycle, to fully
commit its resources. Alternatives should be explored to
reduce that pressure. For example, IDA recipients and
nonborrowers could agree on having a portion of unused
IDA contributions going directly to a trust fund for global
public goods (see below) or rolling them over into the
next cycle. The effects of any changes along these lines
could then be reviewed for subsequent decisions on the
next cycle.
Relations with the IMF in low-income countries
There has been significant discussion within the IMF in
recent years about the changing its role in low-income
countries. With the resolution (for the most part) of the
macroeconomic crises that plagued many low-income
countries in the 1980s and early 1990s, the IMF’s financial
role is likely to diminish over time. Consideration is thus
being given to new modalities for it to monitor and
signal the strength of macroeconomic policies without
direct financial involvement—in the form of “unfunded
CGD0502 0527_Engl_6x9.indd 35
8/17/06 2:55:02 PM
36
Rescuing the World Bank
The Bank
should work
with the
IMF to make
selected IDA
countries
eligible for
automatic
additional
transfers in
the event of
an external
shock
programs” or other “policy support mechanisms.” This
implies that the Bank’s complementary role in judging the
strength of country medium-term development strategies
will become more important—including in assessing for
other donors the appropriate level and composition of
overall support. This will be the case irrespective of the
size—relative to that of other donors—of the Bank’s
financial involvement in a particular country.
This puts a greater onus on the Bank—if it is to maintain
its credibility in countries where it is also providing financial
transfers—to make sufficiently independent judgments
on a country’s policy and institutional status. The new
president should initiate a discussion with shareholders
and the donor community on the criteria for judging a
country’s development strategy and on how the Bank’s
views should most effectively be signaled.
Dealing with external shocks
Many IDA countries are particularly vulnerable to external
shocks, be it weather, a commodity price shock, or a
sudden collapse of the economy of a critical trading
neighbor. In principle, the IMF should help countries
adjust to shocks. But it does not have any grant facility.
The Bank should thus work with the IMF on a facility
to make selected IDA countries at very low income
levels eligible for automatic additional transfers in grant
form. Short-term but rapidly disbursed transfers could
be tied to preselected programs primarily of a social
insurance nature—say, to fund the recurrent costs of a
social insurance nature—say, to fund the recurrent costs
of primary health care. IDA funds would thus be used
to reduce what is otherwise the high and pro-cyclical
volatility of recipient countries’ own revenue and (of even
more volatile and pro-cyclical) overall donor inflows.
Unless and until the IMF can disburse resources in grant
form from its Poverty Reduction and Growth Facility, IDA
resources should be available, with IMF staff technical
input, for this purpose.
For low-income countries, we highlight five specific
recommendations for the new president:
1. Signal support for a narrower, more focused range
of Bank operations within each low-income country,
especially for lending. In the best performing
countries, encourage even more budget support,
keyed to clear benchmarks on results. In poorly
CGD0502 0527_Engl_6x9.indd 36
8/17/06 2:55:02 PM
37
A CGD Working Group Report
performing countries, discourage financial support,
while increasing administrative budget resources for
advisory services, sector work, policy dialogue, and
technical assistance.
2. Urge the shareholders to approve a third, grants-only
window for countries with very low per capita incomes,
for example, below $500.
3. Encourage longer-term commitment periods for the
best-performing countries and programs that
build in more automatic exit when country
performance declines, and propose changes in IDA
replenishment arrangements that would reduce
disbursement pressures.
4. Work with the IMF and other donors and creditors on
an agreed role of the Bank in signaling the adequacy
of a country’s “development” approach to complement
the IMF’s macroeconomic signaling.
5. In collaboration with the IMF, develop a facility to make
selected IDA countries at very low income levels eligible
for automatic additional transfers as grants in the wake
of clearly external shocks.
Take Leadership on Ensuring Truly
Independent Evaluation of the Impact of Bank
and Other Aid-Supported Programs
Agencies that develop and manage development
assistance programs hesitate (with some justification)
to advertise the limits of their craft. The World Bank
is no exception. Although the Bank has improved its
transparency through increased in-house research on
aid effectiveness and through increasingly frank and
systematic work of its internal evaluation department,
neither fills the need for credible, fully independent
assessment.
This is unfortunate. Rigorous and well-targeted
evaluations offer opportunities to substantially expand
the impact of Bank-funded efforts beyond any particular
country or program. The knowledge they generate is itself
a global public good, since the benefit of knowing which
programs work and which do not extends well beyond
the organization or country implementing a program.
Moreover, evaluations of Bank-supported programs
that are fully and visibly independent would improve
CGD0502 0527_Engl_6x9.indd 37
8/17/06 2:55:03 PM
38
Rescuing the World Bank
Evaluations
of Banksupported
programs
that are fully
and visibly
independent would
improve the
credibility of
the Bank’s
efforts
CGD0502 0527_Engl_6x9.indd 38
the credibility of the Bank’s efforts—and that of other
donors—and increase the political support for aid to
support demonstrably effective programs. Independent
evaluation is particularly critical in the IDA countries, for
many of which aid is likely to increase substantially in
the next decade.
In 1973 Bank president Robert McNamara created the
Operations Evaluation Department (OED), a nominally
independent unit within the World Bank reporting directly
to the Bank’s Board of Executive Directors. OED’s
primary mission is to conduct ex post assessments of
Bank-financed interventions. It does this in two ways:
by evaluating projects and by evaluating the Bank’s
development activities more broadly.31 In principle,
OED reports provide analytical background and support
for forward-looking decisionmaking about strategy. In
fact, they are by definition untimely because they are
conducted ex post (often looking back as much as 10
years). And because they are scrutinized in draft by Bank
staff and countries whose programs are the subject of
evaluation, there is a natural process of minimizing the
harshness of language. In addition, it is difficult for even
the best internally sponsored impact evaluations to deal
with such fundamental problems as the lack of baseline
indicators, controls, and a counterfactual.
To address problems of credibility and independence
in evaluation, the Meltzer Commission (International
Financial Institution Advisory Commission 2000), the
Task Force on the Future of the IMF (ODC 2000), and the
Gurría-Volcker Commission (Commission on the Role of
the MDBs in Emerging Markets 2001) all recommended
the creation of an independent evaluation entity external
to the Bank (and the IMF). Gurría-Volcker, for example,
calls on shareholders to create a “mechanism for
independent, third-party evaluation of the effectiveness
for MDB [multilateral development bank] programs [not
just the World Bank], and whether such programs…
encourage adequate norm-setting, increased attention
to poverty reduction, and better policies and stronger
institutions generally.”32
To complicate the challenge, independent evaluation
focused solely on Bank-supported projects can only be
part of the story. As the Bank and other donors move toward
sectorwide and budget support it becomes increasingly
difficult to pinpoint a specific project as being funded
8/17/06 2:55:03 PM
39
A CGD Working Group Report
mainly or entirely by the Bank. But increased country
ownership of investment programs does not reduce the
need for high quality evaluations. Decisions still need
to be made—by governments, local communities, and
others in consultation with the Bank and other donors—
about the best ways to combat poverty, and the Bank
is very well-placed to make an important contribution
to the establishment of an evidence base about the
effectiveness of alternative strategies through rigorous
evaluation. This evidence base can then be drawn upon
The new
by all those involved in development to increase the
president
effectiveness of their programs.
should
The need for impact evaluation of social programs is
lead
the
particularly acute. A forthcoming report of the Center for
creation
of
Global Development will recommend the creation of a
an
external,
voluntary, self-financing consortium of donors, developing
countries, foundations, and international NGOs to sponsor independent,
multidonor
and finance independent impact evaluation of selected
(and
social programs in low- and middle-income countries.
creditor)
aid
The report recommends that some evaluation resources
evaluation
be earmarked for studies with randomized assignment,
which face the largest obstacles relative to their promise mechanism
in knowledge building.33
The Working Group recommends that the new president
lead the creation of an external, independent, multidonor
(and creditor) aid evaluation mechanism to:
Take leadership in working with the board to support
the creation of an independent evaluation entity
financed and governed by a consortium of donors
and multinational creditors.
No one member would have control over the
entity’s operations, but its members would jointly
set priorities about evaluation focus areas. The
reason behind creating a consortium is that a
collective decision, once agreed, would help lock
in good behavior of more and better evaluation—
insulating specific programs from political pressures
associated with negative evaluations.34 This
entity would not focus exclusively on the Bank’s
activities, or even only on donor-financed activities.
It would also assess developing countries’ ownfinanced programs as well as those of NGOs (in
all cases based on requests from these entities).
The consortium could be financed by contributions
CGD0502 0527_Engl_6x9.indd 39
8/17/06 2:55:03 PM
40
Rescuing the World Bank
Past
investments
in global
public goods
relevant to
developing
countries
have had
impressive
rates of
return
CGD0502 0527_Engl_6x9.indd 40
from its individual members, ideally linked to each
member’s own annual aid disbursements.
This entity would assess the effectiveness and
impact of the programs and projects supported by
the Bank and other creditors and donors, not the
policies and processes of the Bank itself (which are
already subject to the Inspection Panel). It would
complement rather than substitute for the audit and
evaluation work of OED (and other internal evaluation
offices of other donor and creditor agencies).
The governance of this entity would be determined
by its members. Ideally developing country members
would join. The Bank’s leadership in creating such
an entity would thus make at least this aspect of
its governance more representative. In any event
decisionmaking for Bank programs would continue
to rest with the board.35
Obtain an Explicit Mandate, an Adequate
Grant Instrument, and a Special Governance
Structure for the Bank’s Work on Global
Public Goods
The last 10–15 years have seen increasing attention to
international initiatives for the financing and provision
of global public goods. Global public goods are those
goods (or “bads”) that no single nation has a sufficient
incentive to produce (or limit) in optimal (from a global
standpoint) amounts, but which have benefits (or costs)
for all nations. Examples include technological advances
in agriculture and health, and global public “bads” such as
global warming. Past investments in global public goods
relevant to developing countries have had impressive rates
of return: as high as 40 percent for agricultural research.36
The return on a malaria vaccine would be comparable.
Investment to reduce or manage expected global warming
would have huge benefits (in reduced economic costs)
that in welfare terms would be greater for developing than
for developed countries.
The Bank has long had some engagement in global
public goods, early on primarily through the Consultative
Group on International Agricultural Research and then
through its role in the Global Environment Facility.
Beginning in the early 1990s, it was drawn into financing
and providing many other, often smaller programs (for
example, support of a consortium of public and private
8/17/06 2:55:03 PM
41
A CGD Working Group Report
agencies working on microfinance issues). These have
generally been housed or run inside the Bank through
specific trust funds financed by interested donors—and
managed outside the purview of overall Bank budget and
program allocations.
The Bank’s status as a global institution with a broad
and deep range of expertise explains the demands from its
shareholders for its technical and financial involvement in
a growing range of global programs, some in the category
of global public goods (and some basically financing
regional and even national programs likely to have some
transnational spillovers). It is now involved (either as a
member, financier, administrator, or participant) in as
many as 70 such programs (table 1). Bank involvement
has helped fill the void created in some areas by U.N.
agencies and the regional development banks’ lack of
comparable financial strength or lack of adequate staffing
and expertise.
But the result is a peculiar situation of the Bank’s
having a set of ad hoc global programs, sometimes
in possible competition with U.N. agencies, without a
clear mandate from its shareholders.The shareholders
have not considered the need for the Bank to have
an instrument comparable to the country loan (with a
sovereign guarantee) that would enable it to pursue such
a mandate strategically—as opposed to responses to ad
hoc requests and ad hoc special financing. As a result,
financing is haphazard. Some programs are financed
from the administrative budget, some from transfers
from net income, and most from Bank-administered
trust funds.37
Without its own instrument, it is difficult for the Bank
to lead in financing or coordinating consortia to finance
new initiatives. As a result, promising programs receive
inadequate attention. A good example is a recent detailed
proposal for an advance commitment to purchase vaccines
for diseases concentrated in low-income countries. The
idea of an advance commitment is to provide incentives for
private firms to undertake the research and development
(R&D) investments needed to develop these vaccines. In
addition, the proposed purchase is structured to ensure
access to these vaccines for the people who need them
most, if and when they are developed. If no vaccine is
developed, no Bank or other donor funds would be spent.
But if successful, millions of lives would be saved at very
CGD0502 0527_Engl_6x9.indd 41
8/17/06 2:55:03 PM
42
Rescuing the World Bank
low cost (for a malaria vaccine, an estimated $15 per year
of life saved).38
The Working Group concluded that beyond the widely
acknowledged objective of reducing world poverty by
supporting equitable growth, there is a case for explicitly
extending the Bank’s mandate for the financing and
provision of global public goods, notably in agriculture,
health, and the environment. Already the Bank, as a key
player in the management of globalization, is seen as a
“go to” institution. But the accretion of responsibilities
The new
without a mandate and clarity on financing and instruments
president
has limited its role and risks ineffective use of its global
should call
resources. We therefore recommend that the new
on the Bank’s president:
shareholders
to give it
a clear
mandate for
financing and
providing
global public
goods
1. Call on the Bank’s shareholders to give it a clear
mandate for financing and providing global public
goods. (We refer here not to any and all forms of global
programs but to those that, because of their public
good nature, have the least call on country-based
financing). Among other benefits, this would give the
Bank clear responsibility for clarifying its contribution
in the light of broader global priorities for investment
in global public goods.
2. Initiate a dialogue with the regional development
banks, the United Nations, and other relevant agencies
on the proper division of labor between global and
regional public goods. In particular, the Bank should
avoid involvement in the latter wherever engagement
by the regional banks makes sense. Between the
banks and the United Nations, there is no obvious
right institutional arrangement that would create
accountability for the financing and implementation
of programs—accountability has to be based on
agreements for respective roles.
3. Ask the board’s members to create a Global Public
Goods Trust Fund managed by the Bank, to consolidate
and help set priorities for current spending from the
Bank’s resources, and to contribute to the financing
of such new and promising initiatives as the advance
market commitment for vaccines.
CGD0502 0527_Engl_6x9.indd 42
8/17/06 2:55:04 PM
43
A CGD Working Group Report
4. Encourage agreement on financing the Global Public
Goods Trust Fund along the following lines:
• Some portion of the Bank’s annual net income
should be earmarked for a Global Public Goods
Trust Fund. The Bank’s net income belongs to all
its members, so its benefits should also extend to
all its members.
• The donor countries that currently make
contributions to the variety of global programs
at the Bank should be urged to contribute instead Some portion
to a single Global Public Goods Trust Fund— of the Bank’s
where they can with other Bank shareholders,
annual net
ensure that priorities for Bank work are aligned
income
with resources, and take into account U.N. and
should be
regional bank activities.
earmarked
• A leaner Bank (thanks to a marked reduction for a Global
in the “hassle” factor and to more automatic Public Goods
and less conditional loans for select eligible
Trust Fund
borrowers) could reduce administrative costs.
The savings could be added to “allocable” net
income and used to supplement the financing of
global public goods. Similarly, more innovative
financial products will induce greater borrowing,
which could increase the net income available for
global public goods.
We estimate that it should be possible, with these
changes, to generate $300–500 million annually for
the Global Public Goods Trust Fund.
5. Encourage agreement on a new approach to the
governance of the Global Public Goods Trust Fund.
Decisions on the use of the trust fund could be made
by the board, but with a different allocation of votes
(akin to the Global Environment Facility, which also
has a different governance structure from the Bank’s
board). IBRD borrowers ought to control at least 40
percent and IDA-only countries another 10–20 percent.
Using net income for global public goods will be
seen by middle-income borrowers as imposing the
costs on them, since their average borrowing rates
would be higher. In particular, to acknowledge their
indirect financing role, the middle-income countries
and emerging market economies should have a seat
CGD0502 0527_Engl_6x9.indd 43
8/17/06 2:55:04 PM
44
Rescuing the World Bank
at the table, with considerable input on decisions on
which global public goods are to be financed.
6. The trust fund rules should clarify that Bank management
need not always be involved in managing the global
public goods that the trust fund helps finance.
No issue
fundamentally affects
the legitimacy of the
Bank—and
its effectiveness—as
much as its
governance
structure
CGD0502 0527_Engl_6x9.indd 44
Push the Bank’s Member Governments
to Make the Bank’s Governance More
Representative and Thus More Legitimate
No issue fundamentally affects the legitimacy of the
Bank—and its effectiveness—as much as its governance
structure. Yet no issue has been as impervious to change.
The Bank should become something closer to the spirit
of a global “club” in which today’s beneficiaries, not
only its rich-country benefactors, have a keen sense of
ownership and financial responsibility.
Votes
Like most major Fortune 500 companies, each of the
agencies that make up the World Bank Group (IBRD,
IDA, IFC, and MIGA) has shareholders that own a stake
in the organization. The one difference, of course, is
that the Bank’s shareholders (unlike most multinational
corporations) are countries rather than individuals. Each
country has a given number of votes linked to the size
of its shareholding.
But the size of country shareholdings no longer
reflects an appropriate balance between borrowers
and nonborrowers. In 1950, for example, when the
countries of Western Europe were the major borrowers
and beneficiaries of the below-market access to
capital the Bank provided, they had some considerable
influence on the Bank’s policies and practices—through
management and staffing as well as their voting shares.
Today, however, the Bank’s borrowers have virtually no
real control over fundamental decisions. For example,
Sub-Saharan African countries represent 27 percent of
all IDA member countries, but have only 8 percent of the
voting shares. Their ownership stake is small, though
they are particularly dependent on the Bank, accounting
for 20 percent of total Bank lending (IDA plus IBRD) in
fiscal 2004.39
Part of the difficulty has to do with the lack of consensus
on when and whether to alter capital shares. Past changes
have come at the time of capital replenishments, when
8/17/06 2:55:04 PM
45
A CGD Working Group Report
the pie was increasing and countries could buy more
shares and increase their percentage of the total. Still,
fast-growing China, now constituting an estimated 13
percent of the world economy, holds just 2.8 percent of
shares, and India, now 6 percent of the world economy
and also growing fast, just 2.8 percent.40 Meanwhile
Saudi Arabia, with 0.6 percent of the world economy,
has 2.8 percent of voting shares (and 1 of the 24 board
seats). Canada and Italy have the same voting shares
as China, and Belgium has 50 percent more votes than
Mexico. In a global club, in any event, other factors,
including population, might ideally affect voting shares
(table 2).41
There is a logic in the continuing power and influence
of nonborrowers. It ensured the Bank’s effectiveness
for many years and it helps sustain their support—in
Table 2
Current and Potential Allocation
of IBRD Voting Power
0.5 (share of population)+
IBRD voting
IBRD voting 0.5 (share of world GDP) (%) share minus (constant
share
Constant
PPP GDP
1995$
(% of total)
1995$ GDP PPP GDP voting sharea billions)
United States
Japan
Germany
France
United Kingdom
Canada
China
India
Italy
Russia
Saudi Arabia
Netherlands
Brazil
Belgium
Spain
Switzerland
Australia
Iran
Venezuela
Mexico
16.4
7.9
4.5
4.3
4.3
2.8
2.8
2.8
2.8
2.8
2.8
2.2
2.1
1.8
1.8
1.7
1.5
1.5
1.3
1.2
15.6
9.3
4.6
3.1
2.4
1.3
12.1
9.2
2.2
1.8
0.4
0.9
2.6
0.5
1.4
0.5
0.9
0.7
0.3
1.4
12.9
4.5
3.0
2.1
2.1
1.2
16.4
11.3
2.0
2.4
0.5
0.6
2.8
0.4
1.2
0.3
0.7
1.0
0.3
1.8
3.5
3.3
1.5
2.2
2.2
1.5
–13.6
–8.5
0.8
0.4
2.3
1.6
–0.7
1.4
0.5
1.4
0.8
0.5
0.9
–0.6
9,196
5,725
2,708
1,832
1,361
741
1,209
517
1,234
469
166
505
810
321
739
339
481
118
75
375
GDP
GDP
(PPP$
billions)
10,357
3,423
2,251
1,604
1,576
960
5,917
2,769
1,529
1,207
273
457
1,352
286
886
218
541
438
134
915
IBRD is International Bank for Reconstruction and Development.
a. A positive value indicates that the current IBRD voting share is too large given population and PPP GDP; a negative
value indidates that a country's IBRD voting share is too small adjusting for population and PPP GDP.
Source: World Bank (2005); IMF (2005b); and author’s calculations.
CGD0502 0527_Engl_6x9.indd 45
8/17/06 2:55:04 PM
46
Rescuing the World Bank
The
continuing
lack of
influence of
borrowers
is likely
to further
undermine
the Bank’s
effectiveness
contrast to their less constant support for many of the
U.N. agencies and the problems of decisionmaking where
the norm is one-country one-vote. Yet in this new century,
more accountable and representative institutions within
countries are seen as more conducive to poverty-reducing
growth, and democracy is broadly acknowledged as the
most legitimate form of government. In this context, the
continuing lack of influence of borrowers reduces the
legitimacy of Bank-supported policies and programs in
some borrowing countries. And over the next decade it
is likely to further undermine the Bank’s effectiveness—
including, ironically, the support for better governance
in borrowing countries.
Voice
The governance deficit is compounded by the inadequate
representation of borrowing countries on the Bank’s board.
Of the 24 board seats, borrowing countries hold only 9;
they share with nonborrowers another 8 (table 3). The
limited representation of developing country borrowers on
the Bank’s board discourages borrowing country board
members from any real scrutiny of other borrowers’
programs. It also creates time and work pressures that
Table 3 Distribution of Voting Power at the Multilateral
Development Banks
Other nonborrower
Developing country
borrowers
United States
Other G-7
Other nonborrower
Developing country
borrowers
Total
President
International
17
Monetary Fund
World Bank
16
Inter-American
30
Development Bank
Asian
13
Development Bank
African
7
Development Bank
European Bank for 10
Reconstruction and
Development
Directors
Other G-7
United States
Voting share (%)
28
17
38
1
6
6
11
24
Nonborrower
27
16
18
4
39
50
1
1
6
4
8
0
9
9
24
14
Nonborrower
Borrower
27
15
45
1
4
1
6
12
Nonborrower
21
12
60
1
4
1
12
18
Borrower
47
30
13
1
6
12
4
23
Nonborrower
Source: Birdsall 2003.
CGD0502 0527_Engl_6x9.indd 46
8/17/06 2:55:05 PM
47
A CGD Working Group Report
make it difficult for them to focus on institutional issues
while also representing their country interests.
The lack of voice in board representation is acute for the
Sub-Saharan countries, which rely heavily on the Bank’s
advice and financial support. At present, 46 Sub-Saharan
countries are represented by just two chairs on the
Bank’s board, creating a tremendous administrative and
procedural burden for the directors and their staffs.
Presidential selection
The president of the Bank is an American, while the
managing director of the IMF is European, under an
implicit post–World War II agreement. This gives the U.S.
administration unchecked discretion in the timing and
process for selecting presidents, undermining the sense of
ownership that ideally would be shared by more member
governments in an institution at the center of a shared
global goal to reduce poverty. The point, however, is not
fundamentally about nationality. It is that the selection
process should be transparent (similar to what the Bank
advocates regarding countries’ governance), and that it
should draw from the global talent pool.42
Role of the board
Another governance problem is the board’s difficulty
in playing a “strategic” role and its inability to make
management accountable to it. For many years, close
observers of the Bank have questioned the effectiveness
of the resident Board, whose members spend full-time
on Bank work and may not have the seniority in their
governments to influence Bank management priorities.
We recommend that the new president engage
early and in open discussion with the Bank governors
on how to address these deficits. If politically difficult
adjustments are to be made, they will almost surely need
to be proposed by the Bank’s president. Despite broad
support from all shareholders for the principle of better
representation, none—and least the most powerful—has
any incentive to make a first move. The result is a deep
problem of collective gridlock. Because the changes will
be difficult, it seems appropriate for the president to open
the discussion during his honeymoon, his first months
in office.
We recommend that the new president take four
specific initiatives to re-establish the legitimacy of the
Bank’s governance:
CGD0502 0527_Engl_6x9.indd 47
8/17/06 2:55:05 PM
48
Rescuing the World Bank
Ask for an
independent
assessment,
to be made
public,
of voting
shares and
board representation,
including
options for
changes.
1. Ask the governors of the Bank to formalize a
credible, rule-based, transparent mechanism (as
with private sector boards) for choosing the Bank’s
president. A 2001 joint report to the Bank and IMF
boards, originated by working groups set up by each
institution, outlined one possible mechanism.43 The
report was endorsed by both boards as guidance
for future selection processes.44 In broad terms, the
report advocated the creation of an advisory group
that would assist the executive directors in presidential
selection by developing a slate of candidates and
providing assessments of each candidate to the
executive directors, who would maintain responsibility
for approving a presidential candidate.
2. Support the temporary establishment (say, for a
decade) of two additional seats on the board for
African countries.45 (In the longer term, if the Board is
to be more strategic, there is a good case for reducing
its size, which could be achieved, for example, with
a decline in the representation of Europe, the most
overrepresented region, and merging the Saudi seat
with that representing other Arab nations).
3. Ask the Bank governors to call for an independent
assessment, to be made public, of voting shares and
board representation, including options for changes.
Options should explore among other issues increases
in the basic votes, the merits of applying double
majorities on some decisions (that is, 50 percent of
all votes plus 50 percent of all members), and should
take into account discussions at the IMF of its quota
distribution during its current quota review period. The
recent communiqué of the International Monetary and
Financial Committee of the IMF on quota reallocations
stated that “adequate voice and participation by all
members should be assured, and the distribution
of quotas should reflect developments in the world
economy.”46 Desirable changes at the IMF and World
Bank would only reinforce what has already been
acknowledged by the creation in 1999 of the Group
of 20 (a new club consisting of the G-8 and such major
emerging market countries as Brazil, China, and India)
to deal with key issues in the international monetary
and financial system.47
CGD0502 0527_Engl_6x9.indd 48
8/17/06 2:55:05 PM
49
A CGD Working Group Report
4. Ask the governors to commission a time-bound
independent review of board functions and
responsibilities. A review of the board should examine
how to make the board more strategic, with emphasis
on its central task of setting objectives and holding
management to account. It could also address how
to trim back the board’s ballooning budget, which
sends the wrong signals on corporate governance.
Meanwhile, push for such interim steps as holding
occasional board meetings in borrowing countries—to
help focus the board on strategic issues in a particular Almost every
region and to foster greater ownership among new regime
borrowers. A board meeting in Pretoria, for instance, enjoys a brief
would highlight the strategic issues of concern to honeymoon
Southern Africa and make it possible to invite particular to put taboo
issues on
borrowing countries to play a more central role in the
the
table for
board meeting, perhaps by giving brief presentations
debate
and
on issues of particular relevance.
We would like to emphasize that the new president’s
agenda on reforming the Bank’s governance structure
is for the medium term. Transforming the Bank from
a traditional development agency to a “club” where
both donors and borrowers have equal ownership and
responsibility will take time, but many Working Group
members considered it the single biggest challenge facing
the new president. Almost every new regime enjoys a brief
honeymoon to put taboo issues on the table for debate
and discussion. A strong statement early on could help
set the tone of the governance debate and give the issue
some much-needed momentum.
discussion
*
*
*
*
*
One temptation the new president should eschew is an
immediate and far-reaching administrative reorganization.
In the past these have been hugely expensive and
disruptive, with little to show for all the smoke and fire.
Instead, we propose two modest changes that would
considerably improve the administrative efficiency
of the Bank. One: simplify regulations to ease out
underperforming staff. For all its bravado about the
CGD0502 0527_Engl_6x9.indd 49
8/17/06 2:55:05 PM
50
Rescuing the World Bank
need for labor market flexibility in its borrowers, the Bank
has been loath to follow its own advice resulting in bloated
costs and lower efficiency. Two: strengthen internal
incentives for staff to work in the poorest and weakest
countries. That would change a common perception that
Bank employees should have significant experience on
the larger, middle-income countries if they are to be
considered qualified candidates for senior positions
within Bank management.
CGD0502 0527_Engl_6x9.indd 50
8/17/06 2:55:05 PM
Biographies of
Working Group Members
K. Y. Amoako
K. Y. Amoako is executive secretary of the Economic
Commission for Africa (UNECA), the regional arm of the
United Nations in Africa, serving at the rank of undersecretary-general of the United Nations. He is also a
member of U.K. Prime Minister Tony Blair’s Commission
for Africa. Prior to joining UNECA in 1995, he served at the
World Bank for two decades and worked in many areas of
the Bank’s activities, including country operations, sector
operations, sector policy, and personnel management.
He also held senior positions, including director of the
Education and Social Policy Department (1993–95).
Owen Barder
Owen Barder is a senior program associate at the Center for
Global Development, where he works on communications
and outreach activities primarily associated with the Policy
Research Network on Global Public Health. Most recently,
he served as the director of Information, Communications,
and Knowledge and head of Africa policy for the U.K.
Department for International Development. He has
also served as private secretary to the Prime Minister
(Economic Affairs) and in the U.K. Treasury, where he
was seconded to the South African Treasury.
Nancy Birdsall
Nancy Birdsall is the founding president of the Center
for Global Development. Before launching the Center,
she served for three years as senior associate and
director of the Economic Reform Project at the Carnegie
Endowment for International Peace. From 1993 to 1998,
she was executive vice president of the Inter-American
Development Bank. Before that she spent 14 years
in research, policy, and management positions at the
World Bank. She is the author, coauthor, or editor of more
than a dozen books and monographs on international
development issues.
51
CGD0502 0527_Engl_6x9.indd 51
8/17/06 2:55:06 PM
52
Rescuing the World Bank
Colin I. Bradford
Colin Bradford is a visiting fellow in economic studies at
the Brookings Institution. He currently serves as adviser
to the Global Economy Track of the Helsinki Process on
Globalization and Democracy. Previously, he was research
professor of economics and international relations at
American University and chief economist of the U.S.
Agency for International Development. He also has held
senior positions at the Organisation for Economic Cooperation and Development, the World Bank, and the Yale
University School of Organization and Management.
Ariel Buira
Ariel Buira is currently director of the G-24 Secretariat.
He is a former staff member and executive director of the
International Monetary Fund. He has been international
director and member of the Board of Governors of
Banco de Mexico, ambassador of Mexico to Greece,
special envoy of the president of Mexico for the U.N.
Conference on Financing for Development, and senior
member of Saint Antony’s College, Oxford. His latest
publications include Challenges to the World Bank and
the IMF (Anthem Press, 2003) and The IMF and the World
Bank at Sixty (Anthem Press, 2005).
Kenneth W. Dam
Ken Dam is Max Pam Professor Emeritus of American
& Foreign Law and Senior Lecturer at the University of
Chicago Law School and senior fellow in Economic Studies
at the Brookings Institution. He has had a distinguished
career in public service, including as deputy secretary in
the U.S. Departments of Treasury and State, executive
director of the Council on Economic Policy, and assistant
director in the Office of Management and Budget. He has
also held positions as vice president for law and external
relations at IBM and as president and chief executive
officer of the United Way of America.
Robert E. Evenson
Robert Evenson is professor of economics and director of
the International and Development Economics Program
at Yale University. He joined the Yale faculty in 1969 and
is the author or co-author of hundreds of scientific papers
on agricultural productivity and economic growth in lowincome countries. From 1997 to 2000 he was director of
the Economic Growth Center at Yale University. His most
CGD0502 0527_Engl_6x9.indd 52
8/17/06 2:55:06 PM
53
A CGD Working Group Report
recent book is Crop Variety Improvement and Its Effect
on Productivity: The Impact of International Agricultural
Research (2003).
Jo Marie Griesgraber
Jo Marie Griesgraber is the director of the New Rules
for Global Finance, a coalition of nongovernmental
organizations and scholars dedicated to the reform of
the global financial architecture. She also serves as vicepresident and secretary of the Financial Policy Forum.
Previously she served as the policy director for Oxfam
America, the director of Rethinking Bretton Woods project
at the Center for Concern, and the deputy director of the
Washington Office on Latin America.
José Angel Gurría
José Angel Gurría is the former minister of foreign affairs
(1994–97) and former minister of finance (1998–2000) of
Mexico and was a Mexican civil servant for 33 years. He
headed Mexico’s foreign financing strategy in the late
1970s and early 1980s and Mexico’s debt restructuring
negotiations during the late 1980s and early 1990s. He cochaired (with Paul Volcker) the Commission on the Role
of the MDBs in Emerging Markets in 2001 and chaired
the External Advisory Group on the future of the InterAmerican Development Bank. He is presently a member of
the board of a number of nonprofit institutions, including
the Center for Global Development, the Population
Council, and the U.N. Secretary General’s Advisory Board
on Water. He is also a member of a number of advisory
boards and boards of directors for private companies in
Mexico, Spain, and the United States.
Pierre Jacquet
Pierre Jacquet has been executive director (in charge of
strategy) and chief economist at Agence Française de
Développement (the French Development Agency) since
2002. He was formerly deputy director of the French
Institute of International Relations in Paris and chief
editor of its quarterly review Politique Etrangère. He is
professor of international economics and chairman of the
Department of Economics and Social Aciences at Ecole
Nationale des Ponts et Chaussées. He is also a member
of the Conseil d’Analyse Economique, an independent
advisory panel created by the French Prime Minister in
July 1997.
CGD0502 0527_Engl_6x9.indd 53
8/17/06 2:55:06 PM
54
Rescuing the World Bank
Edward V.K. (“Kim”) Jaycox
Kim Jaycox is a managing director of EMP Global, a
Washington-based manager of private equity funds
operating in emerging markets. He is also the CEO of
the AIG African Infrastructure Fund. He served at the
World Bank for over 30 years, including as vice president
in charge of the Bank’s operations in Sub-Saharan Africa
from 1984 to 1996. Previously, he directed the World
Bank’s programs in East Asia and led the team that
brought the People’s Republic of China into the Bank.
Devesh Kapur
Devesh Kapur is an associate professor in the
Department of Government at Harvard University.
He serves concurrently as director of the Graduate
Student Associate Program and faculty associate at
the Weatherhead Center for International Affairs and
the Center for International Development at the John F.
Kennedy School of Government at Harvard University.
He is also a non-resident fellow at the Center for Global
Development. He is the author of Give Us Your Best and
Brightest (Center for Global Development, forthcoming)
and co-author of The World Bank: Its First Half Century
(Brookings Institution, 1997).
Michael Kremer
Michael Kremer is Gates Professor of Developing
Societies at Harvard University, senior fellow at the
Brookings Institution, and a non-resident fellow at the
Center for Global Development. He is research associate
at the National Bureau of Economic Research; vicepresident of the Bureau for Research and Economic
Analysis of Development; and a fellow of the Academy
of Arts and Sciences. He serves as associate editor of
the Journal of Development Economics and the Quarterly
Journal of Economics. He is author, most recently, of
Strong Medicine: Creating Incentives for Pharmaceutical
Research on Neglected Diseases (Princeton, 2004).
Steven Radelet
Steven Radelet is a senior fellow at the Center for Global
Development, where he works on issues related to foreign
aid, developing country debt, economic growth, and trade
between rich and poor countries. He was deputy assistant
secretary of the U.S. Treasury for Africa, the Middle East,
and Asia from January 2000 through June 2002. From
1990 to 2000, he was on the faculty of Harvard University,
CGD0502 0527_Engl_6x9.indd 54
8/17/06 2:55:06 PM
55
A CGD Working Group Report
where he was a fellow at the Harvard Institute for
International Development, director of its Macroeconomics
Program, and a lecturer on economics and public
policy. He is the author of Challenging Foreign Aid: A
Policymaker’s Guide to the Millennium Challenge Account
(CGD, 2003).
Jean-Michel Severino
Jean-Michel Severino was appointed director general
of Agence Française de Développement (the French
Development Agency) in 2001. He was previously vicepresident for the Asia region in the World Bank, which
he joined in 1996. Prior to joining the World Bank, he
spent eight years in various positions at the French
Ministry for Co-operation and Development and served
as inspector of finance in the French Ministry of Economy
and Finance. He was nominated general-inspector of
finance and associate professor at the CERDI–University
of Auvergne in 2000.
Vito Tanzi
Vito Tanzi is a consultant at the Inter-American
Development Bank. He spent more than 25 years at
the International Monetary Fund, including as chief of
the Tax Policy Division and director of the Fiscal Affairs
Department. In addition, he has served as undersecretary
in Italy’s Ministry of Economy and Finance, senior
associate at the Carnegie Endowment for International
Peace, president of the International Institute of Public
Finance, and chairman of the Economics Department at
American University in Washington, D.C.
Daniel K. Tarullo
Daniel K. Tarullo is professor of law at Georgetown
University Law Center. From 1993 to 1998 he was,
successively, assistant secretary of state for economic
and business affairs, deputy assistant to the president
for economic policy, and assistant to the president for
international economic policy. From 1995 to 1998 he
was also President Bill Clinton’s personal representative
to the G-7/G-8 group of industrialized nations. Prior to
joining the Clinton administration, he practiced law for
several years in Washington, D.C., mostly in the areas of
antitrust, financial markets, and international transactions.
Previously, he was chief counsel on the staff of Senator
Edward M. Kennedy.
CGD0502 0527_Engl_6x9.indd 55
8/17/06 2:55:06 PM
56
Rescuing the World Bank
John Williamson
John Williamson has been a senior fellow at the Institute
of International Economics since 1981. He was project
director for the U.N. High-Level Panel on Financing for
Development (the Zedillo Report) in 2001 and on leave
as chief economist for South Asia at the World Bank
during 1996–99. He has held faculty positions at Pontificia
Universidade Católica do Rio de Janeiro, University
of Warwick, Massachusetts Institute of Technology,
University of York, and Princeton University. He is
author or editor of numerous studies on international
monetary and developing-world debt issues, including
Delivering on Debt Relief: From IMF Gold to a New Aid
Architecture (Center for Global Development and Institute
for International Economics, 2002).
Ngaire Woods
Ngaire Woods is director of the Global Economic
Governance Programme at University College, Oxford.
She is an adviser to the United Nations Development
Programme’s Human Development Report Office, a
member of the Helsinki Process on global governance,
and a member of the resource group of the U.N.
Secretary-General’s High-Level Commission into Threats,
Challenges and Change. She sits on numerous editorial
and advisory boards, including the Advisory Group of the
Center for Global Development. Her most recent book
is Global Mission: the IMF, the World Bank and Their
Borrowers (Cornell University Press, forthcoming).
Daniel M. Zelikow
Daniel M. Zelikow is a managing director of JP Morgan
and a member of the Government Institutions Group with
responsibility for multilateral financial institutions, export
credit agencies, and some of JPMorgan’s key emerging
markets clients. He also coordinates JP Morgan’s
activities to facilitate Iraq’s financial reconstruction and
helped to found the Trade Bank of Iraq. Before joining
JPMorgan in 1999, he served as deputy assistant
secretary for international affairs at the U.S. Treasury.
Before managing the U.S. financial support program for
Mexico in 1995 as head of the Mexico Task Force, he
directed the Treasury’s overseas technical cooperation,
involving finance ministries and central banks in more
than 20 countries.
CGD0502 0527_Engl_6x9.indd 56
8/17/06 2:55:06 PM
57
A CGD Working Group Report
Notes
1. Einhorn (2001).
2. For a critique from the left, see 50 Years Is Enough
(2004). For a conservative critique, see International
Financial Institution Advisory Commission (2000). For a
useful summary of critiques emanating from both sides,
see Mallaby (2005).
3. See Einhorn (2001).
4. According to an internal Bank audit, “The Bank
faces challenges in effectively customizing its…poverty
reduction strategy to individual countries. The Bank
needs to apply its strategy based on detailed country
knowledge and an appreciation of the willingness and
ability of each country to implement reforms.” See World
Bank, Operations Evaluation Department (2005).
5. For an illustration of this renewed emphasis on
infrastructure, see the final report on IDA-14 replenishment
(IDA 2005).
6. World Bank, Operations Evaluation Department
(2005) suggests the Bank has encouraged too much
lending for social programs in low-income countries,
neglecting the role of infrastructure in growth. It does
not make the point that even among the low-income
group, decisions across programs (agriculture, social,
infrastructure, civil service reform, and more) need to be
made on a country-by-country basis.
7. The Working Group did not discuss the management
question of how much more Bank staff should be
decentralized to work outside of Washington, beyond
the observation that the direction of the last decade
toward greater decentralization has made the Bank more
effective.
8. According to World Bank, Operations Evaluation
Department (2005), the Bank must undertake a “realistic
assessment of the political environment and the
implementation capacity for reform” if it is to strike the
optimal balance between economic growth and long-term
institutional and social development objectives.
9. In turn this requires that the Bank be much more
aware that most of its knowledge generation is for all
practical purposes unavailable to the very audience
CGD0502 0527_Engl_6x9.indd 57
8/17/06 2:55:07 PM
58
Rescuing the World Bank
whose problems it is designed to address—poor and
marginalized communities—because it is rarely in the
languages of these communities. One highly cost-effective
way to increase the impact of the Bank’s expertise is to
make the Bank’s Web site multilingual in the languages
most widely used in client countries. Surveys show strong
demand for this service; yet most of the Bank’s knowledge
is available only in English. See World Bank (2004b).
10. The majority’s vision was that, “all resource
transfers to countries that enjoy capital-market access (as
denoted by an investment-grade international bond rating)
or with a per capita income in excess of $4,000 would
be phased out over the next 5 years.” See International
Financial Institution Advisory Commission (2000), p. 82.
That position had earlier been stated, and has since been
restated and extended, by such distinguished economists
as Kenneth Rogoff, the former director of research of the
IMF, who remarked in a recent public forum that it makes
little sense for the World Bank to be lending to China, with
its high levels of foreign direct investment and growing
dollar reserves, which are the source of huge flows to
the United States.
11. See Commission on the Role of the MDBs in
Emerging Markets (2001). This Commission is often
referred to as the “Gurría-Volcker Commission” after its
chairs, José Angel Gurría and Paul Volcker.
12. The Working Group concluded, in contrast, that
the World Bank should continue to be active in middleincome countries and in such emerging markets as China
and India. Working Group member Daniel Tarullo would be
very cautious about the nature of World Bank involvement
in middle-income and other countries with significant,
sustained inflows of capital.
13. What they lack is long-date, fixed-rate access in
local currency because of investors’ concerns about their
macroeconomic stability.
14. See Commission on the Role of the MDBs in
Emerging Markets (2001), from which some of the text
on this issue is excerpted.
15. To the extent that countries rely solely on access
to private markets (including their own internal markets)
for these investments they are likely to end up with a
dangerous mismatch between short-term liabilities and
long-term returns—thus the problem of vulnerability to
external capital markets.
CGD0502 0527_Engl_6x9.indd 58
8/17/06 2:55:07 PM
59
A CGD Working Group Report
16. Based on Nancy Birdsall and Javed Burki’s personal
discussions and correspondence in 2000–01 with officials
of Brazil, Chile, China, Hungary, India, Mexico, and Poland
as background work for the Gurría-Volcker Commission.
See Commission on the Role of the MDBs in Emerging
Markets (2001). Rodrik (1995) emphasizes that for private
markets, the credibility of the signaling function of the
World Bank and other official creditors rests on the latter’s
view that “in the absence of direct lending, there is very
little to ensure that the official creditors will exercise their
informational function as competently as possible.
17. The objective in principle is for the Bank to cover
non-commercial risk.
18. The Bank might also help out with some risks for
which there is no market (certain commodities, drought)
by owning them directly as an insurer.
19. For one perspective on the proliferation of standards
and their unintended consequences, see chapter 10 of
Mallaby (2004). See also World Bank (2001a, 2001b).
20. This solution was first put forward by Eichengreen
and Hausmann (2003).
21. This approach is proposed in Dervis (2005).
22. This uses the market as a benchmark, with
richer borrowers paying a rate closer to the market rate
they face.
23. One way to do this would be for the Bank to do
more blending of its loans with bilateral grants of donors
into single coordinated operations.
24. See World Bank (forthcoming). For an exposition
specific to Latin America, see Birdsall and de la
Torre (2001).
25. Regarding the Poverty Reduction Strategy Papers,
see World Bank, Operations Evaluation Department
(2004). Regarding reluctance to exit and other failings
of donors in low-income countries, see Nancy Birdsall,
“Seven Deadly Sins: Reflections on Donor Failings,”
Center for Global Development Working Paper Number 50,
December 2004.
26. This is the spirit behind the Bank’s Comprehensive
Development Framework, which by its own description
“emphasizes the interdependence of all elements of
development—social, structural, human, governance,
environmental, economic, and financial.” For more
information on the Comprehensive Development
Framework, see http://www.worldbank.org/cdf.
CGD0502 0527_Engl_6x9.indd 59
8/17/06 2:55:07 PM
60
Rescuing the World Bank
27. See Radelet (2004).
28. Working in these countries is much riskier than
other places. As a result, programs in poorly governed
states require very careful monitoring, regular re-appraisal,
flexible responses as initiatives begin to work or fail, and
a higher tolerance for failure than when working in other
countries.
29. The exact amount would ideally be specified in
purchasing power parity (PPP) terms and in those terms
would probably be higher, as only a handful of countries
are now at an estimated $500 per capita income or less
in PPP terms. In usual exchange rate terms the amount
would ideally be smaller: about 40 countries, including
fast-growing Vietnam, have incomes per capita below
$500 in those terms. On the idea of a second IDA window,
see Radelet (forthcoming).
30. Whether the current three-year cycle of IDA
replenishments affects that ability is not clear. A longer
replenishment period should not be necessary, as IDA
already makes commitments beyond three years. But it
might help, particularly since bilateral aid commitments
over long periods are even more difficult to make.
31. In addition to OED, the Bank created the Inspection
Panel in 1993, a three-member body charged with
providing an independent forum to private citizens
who believe that they or their interests have been or
could be directly harmed by a Bank-financed project.
The Bank’s Executive Board reviews the Panel’s
recommendations and decides whether an investigation
should take place.
32. In the words of the Meltzer Commission
(International Financial Institution Advisory Commission
2000), “The project evaluation process at the World Bank
gets low marks for credibility: wrong criteria combine with
poor timing…The Bank measures results at the moment
of final disbursement of funds. Final disbursement often
occurs more than one year before the project begins full
operations. The start of operations is too early to judge
sustainability of achievements… Evaluation should be a
repetitive process spread over time including many years
after final disbursement of funds, when an operational
history is available” (p. 75). See also Commission on the
Role of the MDBs in Emerging Markets (2001).
33. See
Center
for
Global
Development
(forthcoming).
CGD0502 0527_Engl_6x9.indd 60
8/17/06 2:55:07 PM
61
A CGD Working Group Report
34. Birdsall (2004).
35. Center for Global Development (forthcoming).
36. See Evenson (2003).
37. In fiscal year 2001 (the most recent year for
which data are available), the Bank spent about $30
million of its administrative budget on global programs,
provided another $120 million in grants (also from
its administrative budget under the umbrella of the
“Development Grant Facility”) and disbursed $500 million
from Bank-administered trust funds financed by other
contributors. See World Bank, Operations Evaluation
Department (2002).
38. The U.K. government has proposed supporting, in
collaboration with other donors, such commitments for
malaria and HIV vaccines, and we recommend that the
Bank take a leadership role in supporting this initiative.
The best option would be for the Bank to legally bind
itself to provide IDA loans to any IDA-eligible member that
wanted to purchase the vaccine as long as a number of
pre-specified vaccine characteristics were met. For more
detail on advance purchase commitments, see Center
for Global Development (2005).
39. See World Bank (2004a). The World Bank’s Articles
of Agreement do not allow split voting; all of the votes of
a given “chair” are cast as a unit. As a result, developing
country members of mixed constituencies (for example,
the chairs held by the Netherlands, Belgium, Switzerland,
and Canada) often go unheard on policy matters when
their interests differ from those of the industrial country
that represents them as the chair.
40. Data in this paragraph refer to IBRD voting shares.
Calculations of the shares of world GDP are in purchasing
power parity terms, and data are from the IMF (2005a).
41. Dervis (2005) proposes inclusion of population and
of contributions to the United Nations in his formula for
representation on the U.N. Security Council.
42. The need for transparency in the selection process
was noted by outgoing President James Wolfensohn at
his farewell news conference on May 4, 2005. He referred
to the World Trade Organization model, which recently
chose its new director general from four public candidates.
The recent appointment of a new administrator of the
United Nations Development Programme was also
made following an open selection process, with six
candidates.
CGD0502 0527_Engl_6x9.indd 61
8/17/06 2:55:07 PM
62
Rescuing the World Bank
43. See World Bank and IMF (2001).
44. However, neither board formally adopted the
specific recommendations contained in the report.
45. This echoes a similar recommendation made
by the U.K.-sponsored Commission for Africa (2005),
which advocates for two new African chairs on the
Boards of the World Bank and IMF: “As the rules for
representation on the Boards [of the World Bank and
IMF] are based on economic criteria, it is not likely that
African representation will exceed two chairs out of 24
in the short term. However, a decision could be taken by
consensus to allow the creation, on a temporary basis
(for the entire period up to 2015), of two supplementary
positions of Executive Director for Africa, each backed
by an Alternate Director, in each Board. This would ease
the task of the directors in this critical period for Africa’s
development” (p. 368).
46. The communiqué states: “The IMF’s effectiveness
and credibility as a cooperative institution must be
safeguarded and further enhanced. Adequate voice and
participation by all members should be assured, and
the distribution of quotas should reflect developments
in the world economy. The Committee emphasizes that
the period of the Thirteenth General Review of Quotas
provides an opportunity for the membership to make
progress toward a consensus on the issues of quotas,
voice, and participation” (IMF 2005b).
47. See also Dervis (2005).
References
50 Years Is Enough. 2004. “10 Things You Really Should
Know about the World Bank.” Washington, D.C.
Birdsall, Nancy. 2003. “Why It Matters Who Runs the
IMF and the World Bank.” Working Paper 22. Center for
Global Development, Washington, D.C. Also in Josack, S.,
G. Ranis, and J. Vreeland, eds. Forthcoming. Globalization
and the Nation State: The Impact of the IMF and the World
Bank. London: Routledge.
CGD0502 0527_Engl_6x9.indd 62
8/17/06 2:55:07 PM
63
A CGD Working Group Report
Birdsall, Nancy. 2004. “Seven Deadly Sins: Reflections
on Donor Failings.” Working Paper 50. Center for Global
Development, Washington, D.C.
Birdsall, Nancy, and Augusto de la Torre. 2001.
Washington Contentious: Economic Policies for
Social Equity in Latin America. Washington, D.C.:
Carnegie Endowment for International Peace and the
Inter-American Dialogue.
Center for Global Development. 2005. Making Markets
for Vaccines: Ideas to Action. Report of the Advance
Market Commitment Working Group. Washington, D.C.
Center for Global Development. Forthcoming. “When
Will We Ever Learn? Recommendations to Improve Social
Development Assistance through Improved Impact
Evaluation.” Report of the Development Evaluation Gap
Working Group. Washington, D.C.
Commission for Africa. 2005. Our Common Interest.
London.
Commission on the Role of the MDBs in Emerging
Markets. 2001. The Role of the Multilateral Development
Banks in Emerging Market Economies. Carnegie
Endowment for International Peace, EMP Financial
Advisors, LLC, and the Inter-American Dialogue,
Washington, D.C.
Dervis, Kemal. 2005. A Better Globalization: Legitimacy,
Governance, and Reform. Washington, D.C.: Center for
Global Development.
Eichengreen, Barry, and Ricardo Hausmann. 2003. “The
Road to Redemption.” In Other People’s Money: Debt
Denomination and Financial Instability in Emerging-Market
Economies. Chicago, Ill.: University of Chicago Press.
Einhorn, Jessica. 2001. “The World Bank’s Mission
Creep.” Foreign Affairs 80(5): 22–31.
Evenson, Robert. 2003. “Production Impacts of
Crop Genetic Improvement.” In R. E. Evenson and D.
Gollin, eds., Crop Variety Improvement and its Effect on
Productivity. Wallingford, U.K.: CABI Publishing.
IDA (International Development Association). 2005.
“Additions to IDA Resources: Fourteenth Replenishment—
Working Together to Achieve the Millennium Development
Goals.” Report from the Executive Directors of the
International Development Association to the Board of
Governors. Washington, D.C.
IMF (International Monetary Fund). 2005a. “Communiqué
of the International Monetary and Financial Committee
CGD0502 0527_Engl_6x9.indd 63
8/17/06 2:55:08 PM
64
Rescuing the World Bank
of the Board of Governors of the International Monetary
Fund.” Press Release 05/87, April 16. Washington, D.C.
———.
2005b.
World
Economic
Outlook.
Washington, D.C.
International Financial Institution Advisory Commission.
2000. “International Financial Institutions Reform.”
Washington, D.C.
Mallaby, Sebastian. 2004. The World’s Banker: A Story
of Failed States, Financial Crises, and the Wealth and
Poverty of Nations. New York: Penguin Press.
———. 2005. “Saving the World Bank.” Foreign Affairs
84(3): 75–85.
ODC (Overseas Development Council). 2000.
“Report of the Task Force on the Future of the IMF.”
Washington, D.C.
Radelet, Steven. 2004. “Aid Effectiveness and the
Millennium Development Goals.” Working Paper 39,
Center for Global Development, Washington, D.C.
———. Forthcoming. “Grants or Loans? How
Should the World Bank Distribute Funds to the World’s
Poorest Countries?” Center for Global Development,
Washington, D.C.
Rodrik, Dani. 1995. Why Is There Multilateral Lending?
NBER Working Paper 5160. Cambridge, Mass.: National
Bureau of Economic Research.
World Bank. 2001a. “Report of the Task Force on the
Cost of Doing Business.” Washington, D.C.
———. 2001b. “Report of the Task Force on the
World Bank Group and the Middle-Income Countries.”
Washington, D.C.
———. 2004a. Annual Report 2004: Volume One.
Washington, D.C.
———. 2004b. “Why The World Bank Needs a
Multilingual Web Site and What to Do About It.”
Washington, D.C.
———. 2005. World Development Indicators.
Washington, D.C.
———. Forthcoming. World Development Report 2006:
Equity and Development. Washington, D.C.
World Bank, Operations Evaluation Department.
2002. The World Bank’s Approach to Global
Programs—An Independent Evaluation: Phase 1 Report.
Washington, D.C.
CGD0502 0527_Engl_6x9.indd 64
8/17/06 2:55:08 PM
65
A CGD Working Group Report
———. 2004. The Poverty Reduction Strategy Initiative:
An Independent Evaluation of the World Bank’s Support
Through 2003. Washington, D.C.
———. 2005. 2004 Annual Review of Development
Effectiveness: The Bank’s Contributions to Poverty
Reduction. Washington, D.C.
World Bank and IMF (International Monetary Fund).
2001. “Draft Joint Report of the Bank Working Group to
Review the Process for Selection of the President and The
Fund Working Group to Review the Process for Selection
of the Managing Director.” Washington, D.C.
CGD0502 0527_Engl_6x9.indd 65
8/17/06 2:55:08 PM
CGD0502 0527_Engl_6x9.indd 66
8/17/06 2:55:08 PM
SELECTED
ESSAYS
CGD0502 0527_Engl_6x9.indd 67
8/17/06 2:55:08 PM
CGD0502 0527_Engl_6x9.indd 68
8/17/06 2:55:08 PM
A Global Credit Club, Not Another
Development Agency
by Nancy Birdsall
In 2000 the majority report of the Meltzer Commission1
called for the World Bank to get out of lending and move
to grants and small technical-assistance programs for
the poorer countries—to become a “World Development
Agency.” In one of the essays in this volume, Adam
Lerrick makes a similar case—that because private
capital is now available to many developing countries,
and given that many developing countries are indeed
borrowing less from the World Bank, it is time to push
the shareholders and management to focus much more
on the poorest countries.
In this essay I argue that the last thing the world needs
is another development agency. We have a multitude
of those—USAID, the British DFID, and the bilateral aid
agencies of at least two dozen other advanced economies;
UNICEF, UNIFEM, UNDP, and the European Union; the Red
Cross, Oxfam, and World Wildlife Fund; and so on.2 What
the world does need is more global clubs of countries—
where decisions, as in country-based democracies,
are based on shared discourse, and implementation of
decisions is effective because the process is viewed as
legitimate in reconciling conflicting views. (The word
“club” has different connotations in different cultures
and settings. I use it in the everyday “American” sense,
which implies open membership not exclusivity—for
example the local Rotary Club not the country club.)
The World Bank can be thought of as a particular type
of global club, with a structure close to that of a credit
union in which the members are nations. Its mission
is in the common interests of all its country members:
broadly shared and sustainable global prosperity.
Nancy Birdsall, Center for Global Development
Nancy Birdsall is the founding president of the Center for Global
Development. Before launching the Center, she was the senior associate
and director of the Economic Reform Project at the Carnegie Endowment for
International Peace. Birdsall was previously executive vice president of the
Inter-American Development Bank and director of the World Bank’s research
department. She is the author, co-author, or editor of more than a dozen
books and monographs.
CGD0502 0527_Engl_6x9.indd 69
69
8/17/06 2:55:08 PM
70
Rescuing the World Bank
(Economists might think of this mission as one of working
for convergence—of accelerating the process by which
relatively poorer nations converge, through development
and transformation, toward the prosperity of their richer
counterparts.)
In the light of this (simple) idea of the Bank as a credit
club, I review here the challenges the Bank faces, including
those set out in the preceding report as “five crucial tasks”
for the Bank’s newest president (in the report I co-chaired
with Devesh Kapur, hereafter referred to as the“Working
Group Report”), and those debated and discussed in the
11 essays that follow.
Bretton Woods: Inventing a global credit club
The World Bank is not of course the only global club
(the largest in number of members is obviously the
United Nations), and it is not the only credit union whose
members are countries—there are, for example, the
regional development banks, the European Investment
Bank, and for some aspects of development there is the
International Monetary Fund. However, it is the only truly
global club that has the financial structure of a credit
union. Let us call it, informally, a “global credit club.”
In this global credit club, different members have
different amounts of “deposits” and provide different
amounts of guarantees. The biggest depositor is the
U.S. government and, along with Europe and Japan, the
United States is the World Bank’s biggest guarantor. It
and its rich country colleagues back all the borrowing of
this peculiar credit union, whether the credit union makes
good loans or bad loans, and whether its member country
borrowers pay up or not (though history indicates that only
rarely do they fail to pay on time).3 The guarantees (and
perhaps the extraordinarily low default rate) mean that
this credit union, even with relatively low deposits in the
form of paid-in capital, can borrow outside at good rates,
and lend at good rates to its less wealthy members.
The global credit club was the brilliant invention of
U.K. economist John Maynard Keynes, along with the
American, Harry Dexter White, and their colleagues
from 42 other countries who conceived the Bank and
the IMF at the Bretton Woods Conference in 1944. They
conceived of a global club which, at low financial cost to
the big depositors and guarantors (at that time, only the
United States for all practical purposes), would reduce
CGD0502 0527_Engl_6x9.indd 70
8/17/06 2:55:09 PM
71
Selected Essays
borrowing costs for the poorer members (at that time
war-torn Europe) and make the world richer and safer.
The boundaries within which the club would operate
were well understood and fully embraced by the club
members. This club was established to promote an
open and liberal international economic system, based
on market-driven growth and trade (in notable contrast
to the system espoused by the Soviet Union (then in
1944 a wartime ally). It would do so by helping the warravaged countries of Europe and the poorer countries of
Asia, Latin America and Africa make the investments that
would enable them to prosper as partners in this open
system—in the interests of global stability and security.
Note the financing mechanism for this global credit club
did not rely on “contributions” to finance “transfers” from
rich to poor nations (though later the club members created
a separate club for that purpose, called IDA, in which only
the rich country contributors have membership rights).
Keynes and his colleagues did not invent a development
agency, and did not conceive of the resulting financing as
a transfer. On the contrary, the borrowers (at that time to
be primarily the Western Europeans) were thought of as
full members and partners in the club’s venture. Today,
as is the case with an everyday credit union, the World
Bank’s capacity to make loans at low costs to borrowers
arises because the sum of the membership’s credibility
reduces borrowing costs for all members below what
they would pay on their own. (This would be true even
for the rich countries—their cost of borrowing would be
reduced slightly because of the lower risk associated
with their diversity. Even today, Germany borrows from
the European Investment Bank.)
Today: An aid agency?
It is surprising how far the World Bank of today has
strayed, in spirit at least, from this original conception.
To quote Jessica Einhorn in her recent essay on the World
Bank in Foreign Affairs: “Over time, the Bank has evolved
from an organization focused on growth through trade
and investment to an organization set on achieving a
world without poverty. Its core mission is no longer to
partner with … countries in their pursuit of balanced and
externally oriented growth; it is to alleviate poverty…”
(italics added).4 And to quote from the Working Group
CGD0502 0527_Engl_6x9.indd 71
8/17/06 2:55:09 PM
72
Rescuing the World Bank
Report: “The Bank’s mission is to reduce poverty in
developing countries.”5
Much of the discussion and debate about the World
Bank today—its effectiveness, its relevance, and its
legitimacy—is framed by this different way of describing
its basic mission. The different conception is not entirely
recent. The non-borrowers established the IDA window
for lending to the poorest countries on the basis of outright
contributions in 1960—creating in effect an “aid agency”
inside the existing club. The speech of Robert McNamara,
the Bank’s fifth president, in Nairobi in 1973 perhaps
marks the official birth of the “poverty” mission for the
Bank group as a whole, including the IBRD. Up to that
time the Bank was primarily a financier of bricks and
mortar projects, with investment in infrastructure seen
as the key to open, market-based growth. By the time of
James Wolfensohn, the poverty objective had matured
and was captured aptly in the lobby of the Bank’s main
building: “Our dream is a world free of poverty,” and in
a noteworthy increase in the proportion of Bank lending
for social programs.
Is there much real difference between a credit club with
an objective of shared global prosperity in an open liberal
economy, and a development agency to battle poverty,
given that market-led growth and poverty reduction are
generally mutually reinforcing? The difference is in part
between a club with a mission in every member’s interest
(global security and prosperity in an open system), versus
an aid agency in which some parties are “contributing”
to further the interests of others. But in terms of what
the Bank is meant to do, broadly defined—provide
loans to help countries accelerate their growth and
development (and reduce poverty)—there is of course
no obvious difference.
Where differences do arise, however, is in the specific
priorities and choices and the process for making those
choices. Thus over the last several decades, the debate
about the relative importance of “growth” versus “poverty
reduction” at the Bank has been associated with periods of
emphasis on lending for infrastructure for “growth,” versus
health and education for “poverty reduction.”6 Perhaps
more in the spirit of an aid agency than a cooperative,
the pendulum is now swinging back to “infrastructure,”
but with the explicit objective of “poverty reduction”!
Similarly, recently the pendulum has swung away from
CGD0502 0527_Engl_6x9.indd 72
8/17/06 2:55:09 PM
73
Selected Essays
“conditionality” (a process associated with the Bank and
often its powerful non-borrowers insisting on their view of
what policies would generate development) to the more
club-like spirit embodied in the emphasis on “ownership”
by member borrowers of their own reforms, and on the
importance of “participation” within developing countries
of citizens in deciding on reforms.7
Three problems, five tasks
The fact is that the distinction between club and aid
agency, subtle as it may be, matters for the future of the
World Bank. The Bank is under tremendous pressure
today. It is assailed from the left for lack of legitimacy—for
promoting privileged “insider” financial and corporate
interests instead of addressing the needs of the voiceless
poor. It is assailed from the right for its refusal to admit to
its lost relevance; with increasing flows of private capital
to the developing world (and ample reserves in China,
India and many other emerging markets), why use public
resources to subsidize loans to those settings—where
private markets and private transfers would be more
efficient and effective?8 In the center, from inside as well
as outside the Bank, it is criticized for lack of effectiveness
in attacking poverty in the poorest countries, for its lack
of agility in responding to the real demands of its largeand middle-income borrowers (and thus its apparent
loss of relevance), and for its loss of institutional focus,
as it responds to ever-expanding demands on it from
(ironically some would say) its more powerful members:
to do everything from assessing needs in Gaza and Iraq,
to managing a global program to “fast-track” education
gains, to piloting trading across countries in carbon
emissions rights.
The pressures have to do with three problems—erosion
of the Bank’s legitimacy as an institution, loss of faith in
its effectiveness (in reducing poverty and in promoting
“balanced and externally oriented growth”), and its
apparent growing irrelevance. It is these problems in their
various forms that authors of the preceding Working Group
Report and the essays that follow—all supporters of the
Bank’s fundamental mission and of its continued existence
in some form—address, with proposals for change and
reform.
How might the Bank’s shareholders, especially the
United States, by embracing a vision of the Bank as a
CGD0502 0527_Engl_6x9.indd 73
8/17/06 2:55:09 PM
74
Rescuing the World Bank
club (rather than as a development agency) be better
positioned to address these problems? How might a
return to the spirit of Bretton Woods, to the idea of a
global credit club, change the outlines of the current
debates among the Bank’s critics about the institution?
How might that conception shape changes in the policies
and practices of the Bank, including along the lines of the
proposed “five crucial tasks” set out in the report above?
I address the “club” question now in the context of the
five issues or tasks set out in that report.
1. Governance: Does the Bank have legitimacy?
How did the founders make operational the idea of a global
credit club? They agreed that in this club, voting power
would be related to members’ “dues” (or deposits and
guarantees), and the deposits and guarantees would be
broadly related to members’ financial capacity. However,
they were concerned to avoid a perfect one-to-one
relationship between financial capacity and influence in
the club. On the one hand, members taking greater risk
ought to have substantial say in the rules and practices
of the club—if only to secure their continued financial
commitment. On the other hand, the overwhelming
financial capacity of a very few countries to take that
risk, if reflected fully in the allocation of votes, would
undermine the spirit of a club. As Harry Dexter White
noted at the time (referring to the International Monetary
Fund), “to accord voting power strictly proportionate
to the value of the subscription would give the one or
two powers control over the Fund. To do that would
destroy the truly international character of the Fund,
and seriously jeopardize its success.” 9 Therefore in the
case of the Bank and the IMF the founders introduced
such mechanisms as “basic votes” that were distributed
equally to all members (in the Bank each member has
250 votes irrespective of shares, plus one additional vote
for each share), and double majority voting (of shares
and of member countries) to make certain fundamental
changes in the Articles of Agreement.
The idea was that the country taking the main risk—
at that time the United States—would define the key
boundaries within which the club’s operations would
work. At the same time, to preserve the spirit of a club and
to ensure that the club would be effective, other members,
including active borrowers (initially the Europeans) would
CGD0502 0527_Engl_6x9.indd 74
8/17/06 2:55:09 PM
75
Selected Essays
have opportunities to influence, within those boundaries,
the club’s specific priorities, policies and detailed
practices, and on some key issues, the ability to resist
changes that might reflect only the narrow interests of a
few powerful members.
Over time, however, whatever ability and interest the
Bank’s initial mostly European borrowers had to affect
the Bank’s priorities, policies and practices have clearly
eroded for today’s many more numerous borrowers.
In 1947–1948 the Bank made loans to six countries
(France, the Netherlands, Denmark, Luxembourg, China
and India). Today the IBRD and IDA lend to almost
150 countries. And the world has changed in another
respect. Political mechanisms of representation and
voice in “democracies” and in international “clubs” of
nations are now almost universally acknowledged as
ideal in their own right (Development as Freedom, to
use the title of Amartya Sen’s book), and as effective
in an instrumental sense—for sustainable growth and
poverty reduction because they create accountability and
produce checks on abuse of power. The idea of political
freedom in a democracy is also now closely associated
with the Western economic model of open markets, and
thus with the original “mission” of the club. International
clubs are not immune from these changes in norms.
The result, reflected in the report and essays in this
book, is a growing demand for reform of governance at the
Bank, especially to ensure much greater representation—
in terms of voting power, Board membership, staffing,
and so on—of developing country borrowers who are the
members most affected by Bank policies and practices.
The spirit of an international club is particularly resonant
in the proposals for:
• The current president of the Bank to “push the
Bank’s member governments to make the Bank’s
governance more representative and thus more
legitimate” and commit now to a more open and
transparent process for selection of his successor
(Working Group Report).
• Use of double majority voting on many more issues
to create an incentive for borrowers who now
see no point in debating institutional issues over
which they have no influence to build coalitions
(Ngaire Woods).
CGD0502 0527_Engl_6x9.indd 75
8/17/06 2:55:09 PM
76
Rescuing the World Bank
• A governance structure for a trust fund for global
public goods at the Bank in which the middleand low-income borrowers would have at least
50 percent of the votes, with the middle-income
countries having more power to set the agenda in
return for the financing they would be providing by
paying higher interest charges on their loans than
otherwise (Working Group Report).
• A rethinking of the “framework” for the IDA window,
separate from any reconfiguration of IBRD shares
(which would have little impact on decision-making
in the IDA), so that both donors and recipients would
“feel more ownership” (Masood Ahmed).
With a more representative governance structure
and broader engagement of borrowers, the Bank would
obviously look more like a club, and looking more like
a club would command more legitimacy as a global
institution. It would still be a credit club, in which the big
depositors have more say. But it would also provide much
greater incentives for borrowing members to engage on
key issues. To quote Harry Dexter White once again:
“Indeed it is very doubtful if many countries would be
willing to participate in an international organization with
wide powers if one or two countries were able to control
its policies.” 10
2. The low-income countries: Is the Bank effective?
Masood Ahmed in his essay makes the point that without
reform of its governance, the Bank risks additional
erosion not only of its legitimacy but its effectiveness:
“I am persuaded that the World Bank cannot continue
to deliver the results we all want over the next decade
without substantial governance reform.”11 Ngaire Woods
also links the Bank’s problems of effectiveness to its
inadequate governance: “current arrangements have
proven to be ineffective from a corporate governance
point of view as well as from a political ‘legitimacy’ point
of view.”
Ahmed is referring primarily to the Bank’s efforts to end
poverty, particularly in the poorest, most aid-dependent
countries. On that effort, William Easterly goes further,
arguing that the Bank is already failing to deliver results—
failing because it lacks mechanisms of accountability
for results to the poor people whom it is meant to be
helping. For Easterly, lack of accountability is rooted in
CGD0502 0527_Engl_6x9.indd 76
8/17/06 2:55:10 PM
77
Selected Essays
the lack of political voice of the poor—of their countries
in the Bank, and in many cases of the poor in their own
countries.12 Steven Radelet is equally critical of the Bank’s
effectiveness in the poorest countries most dependent
on the Bank, saying that the Bank encourages recipient
governments to “take on way too many issues and
activities, leading to no focus, no sense of priorities, and
less progress on really key issues.”
The Bank, along with virtually all official creditors and
donors, is now committed to the principle of “ownership”
by developing countries of their own policies and reforms.
Yet as long as the Bank itself is not seen as “owned”
and legitimate in developing countries, it is too easy for
Bank-financed programs to become controversial and
difficult for developing country leaders to implement.
The plight of Bank adjustment programs (and of the
much benighted “Washington Consensus” reforms in
general) is a compelling example; Easterly (2002) among
others documents their failure in many low- and middleincome countries. Bank programs become controversial
not only because they have losers as well as winners
(which cannot be avoided), but because they are seen
as imposed from outside.
Returning to the spirit of Bretton Woods might help. In a
club (but not in an aid agency), the recipients of financing
have the power that comes with membership, and their
agreement is more obviously required on the broad policies
and practices that govern the financing process.
3. China, India and the middle-income countries:
Is the Bank still relevant?
It could be argued that since the IDA window for the lowincome countries is financed by contributions from the
rich countries, IDA is the “aid agency” of the Bank. But
that is not the case with the IBRD window, the source of
financing for China, India, and most of the world’s “middleincome” countries (in World Bank parlance, countries with
income above $825 per capita). The club-like nature of
the Bank rests largely on its IBRD functions.
Adam Lerrick argues persuasively that the Bank should
shift its resources from loans for middle-income countries
to grants for the poorer countries, on the grounds of
the “irrelevance of lending” in a world of sophisticated
private markets. He and others point to the decline in
borrowing and the acceleration of loan repayments by
CGD0502 0527_Engl_6x9.indd 77
8/17/06 2:55:10 PM
78
Rescuing the World Bank
many middle-income countries in recent years. On at
least that score, for at least one country right now, he has
a point: “China is awash in money.”13 David de Ferranti
answers, equally persuasively, noting the volatility of
markets, the poor access of some of the smaller and
poorer “middle-income” countries and the broad-based
analytical knowledge the Bank brings to issues of global
importance, such as financial crisis prevention and
environmental protection. The Working Group Report
adds the argument of the legitimate interest of the rich
country non-borrowers in promoting equitable growth
in countries where two-thirds of the world’s poor live,
including in support of their own prosperity and security
(pp. 21–22 in this text).14
De Ferranti and the Working Group Report make various
proposals for retaining the allegiance of middle-income
borrowers (and thus retaining access to the net income
their borrowing generates), emphasizing the need both
for new products to catalyze private flows to countries,
and for reduced costs of the “hassle” associated with
borrowing from the Bank, be it due to: excessive (or not)
fiduciary obligations including to limit corruption, excessive
(or not) safeguards against environmental and other costs,
or the political and financial costs of excessive (or not)
delays between requesting and receiving a loan . Similarly,
Jessica Einhorn (2006) suggests that the Bank’s members
agree to lock in now a 25-year sunset clause for loan
disbursements, as an incentive for the Bank management
and bureaucracy to adapt itself to the creative challenge
of developing a new set of non-loan services for middleincome countries more quickly.
If we conceive of the Bank as a club, managed by
its members for their own benefit, then the substantive
merits of these arguments and proposals for change, one
way or another, yield to the question of whether particular
members wish or not to avail themselves of the benefits
their membership provides—under existing conditions—
and/or use their influence to change the conditions.
If China wants to borrow at the cost already agreed
to by all the members, for whatever reason (including
because China trusts more the technical input of the
Bank if it is bundled with a financial commitment), then
so be it. If a country (Korea in 1998) that had eschewed
borrowing for many years asks the Bank for a loan
during an emergency, then so be it. If non-borrowers
CGD0502 0527_Engl_6x9.indd 78
8/17/06 2:55:10 PM
79
Selected Essays
wish to limit the subsidy implied in loans to relatively
rich or more liquid (in terms of reserves) middle-income
members, then they have the option of proposing a
policy of smaller subsidies (higher interest charges on
loans) for the relatively rich borrowers.15
Put another way, let the members of the club decide.
In effect that is the current situation—though it reflects
as much the inertia of failed cooperation as a positive
decision. An interesting issue arises because some
members, and particularly the borrowers, have limited
influence in changing the conditions (pricing, delays,
conditionality, and safeguards) under which they now
participate as members. In that sense, the governance
question—whether the Bank can return to its roots as a
global club—is key to whether it continues to be relevant
in its current form for a large group of its members. In the
absence of voice, some members may in effect choose
the option of exit.16
4. Global public goods and independent
evaluation: Is the World Bank a “knowledge
bank?” Would a “knowledge bank” be
more relevant?
The Working Group Report suggests the Bank is
uniquely positioned for greater strategic involvement
in the production and financing of global public goods.
It is so positioned both because of its potential to
finance production of such goods (including by others),
and because of its combination of a global “macro”
perspective on the costs and benefits of such goods
with specific technical expertise in relevant sectors, such
as agriculture, health and environment. Michael Kremer
provides four compelling examples where the Bank ought
to be active: health and agricultural technologies for the
poor; an African road network (co-financed with the African
Development Bank); financial support for countries that
take in and integrate refugees; and the development of
global knowledge on the impact of various public policies
in developing countries.
Certainly, greater involvement of the Bank in global
public goods, with agreement on priorities by the
members, makes sense. (At the moment, the Bank does
have programs in global goods, but they are financed
and managed in an ad hoc way, often relying on special
contributions from one or two rich country members.) A
CGD0502 0527_Engl_6x9.indd 79
8/17/06 2:55:10 PM
80
Rescuing the World Bank
program of support for global public goods would enliven
the “club” spirit at the Bank. That would be particularly
so were it supported, as recommended in the Working
Group Report, through a large, new trust fund financed
by direct contributions from members (presumably mostly
non-borrowers), and by pre-agreed annual transfers from
net income due to the Bank’s loans—implying indirect
financing by Bank borrowers. A separate financing and
governance arrangement would in effect constitute a new
club within the existing club.
The question, however, is whether the Bank as an
institution does currently gather and convey useful
“knowledge” on development practice. On the one
hand, supporters of the Bank increasingly invoke its
comparative advantage in generating and disseminating
worldwide knowledge and expertise on development
policy and practice. The Working Group Report, for
example, refers to the Bank as “development’s brain
trust” and as a global public good (pp. 17–18).17 On the
other hand, others are deeply critical. Devesh Kapur
asks whether the Bank’s spending on the production
and dissemination of knowledge is cost-effective relative
to direct spending on other global public goods: “If the
Bank’s overall budget was cut by a third and the resulting
savings (more than one half billion dollars annually) were
put into research in those diseases, crops and energy
technologies that are sui generis to poor countries, would
the global welfare of the poor improve or decline?” He
links the virtual absence of Bank reliance on researchers
based in developing countries to Bank researchers’
greater interest in “propositional” knowledge—the search
for universal laws of development—and “prescriptive”
knowledge, rather than in “a deeply textured knowledge
of the circumstances” of a country that could provide
guidance on how to build institutions and who might
do so.
Levine and Savedoff worry, in a similar vein, that “the
Bank rarely creates new knowledge about what works.”
They describe a track record that is “wanting” on the very
sort of program that Kremer calls on the Bank to support:
impact evaluation of programs in developing countries.
To address the effectiveness of the Bank as a knowledge
bank, Levine and Savedoff call on the Bank to encourage
and support much more impact evaluation, including of
the programs and projects it finances, and to join in a
CGD0502 0527_Engl_6x9.indd 80
8/17/06 2:55:10 PM
81
Selected Essays
collective response to ensure “supply of knowledge, a
global public good in the truest sense.” Kapur wants
more outsourcing of research and knowledge creation
to scholars based in developing countries, and more
emphasis on the production of country-specific
knowledge. Kremer calls for more direct financing of
research in agriculture and health likely to have global
benefits, presumably including research done primarily
outside the Bank. Pierre Jacquet calls on all donors and
creditors, including the Bank, to agree on benchmarking
of their programs against results of evaluations.
But (as Kapur notes) there are now no incentives for the
Bank, as a bureaucracy, to outsource research. Indeed,
short of specific contributions for specific programs, the
Bank as a bureaucracy has no incentive to “do” global
public goods beyond its own in-house “knowledge”
activities. And as Levine and Savedoff note, there are
disincentives for Bank staff to promote impact evaluation
of Bank-financed programs. Under current conditions, it
is not clear that the Bank can be an effective “knowledge”
bank, even regarding learning about its own effectiveness.
(Nor is it clear that under current bureaucratic incentives,
Easterly’s and Radelet’s worries that the Bank is not
accountable for results and not able to set priorities in
low-income countries, or the complaints of others that
the Bank creates too much hassle for middle-income
borrowers, will be addressed.)
But perhaps proposals to exploit and to fix the Bank
as a “knowledge” institution could be realized in a more
club-like environment. The development literature is
now replete with invocations of the simple reality that
developing countries are ultimately responsible for their
own fates. Accountability for results of development
efforts must rely ultimately on the political mechanisms
by which governments are accountable to their citizens
and by which international institutions are accountable
to their members. A club might more obviously create
accountability of its staff to all its members.
To argue that the Bank explicitly recognize its potential
comparative advantage as a “club” is not to suggest
that it become less businesslike. Indeed, an alternative
metaphor for a more effective and relevant Bank, that
of a competitive firm subject to market discipline, leads
to much the same conclusions about the need for
reform. Mark Stoleson, of a global private investment
CGD0502 0527_Engl_6x9.indd 81
8/17/06 2:55:11 PM
82
Rescuing the World Bank
firm, makes this point with unusual freshness and clarity
in the final essay of the volume “The World Bank: Buy,
Sell or Hold.”
A concluding note
In any event, it is an over-simplification to call the Bank a
club. Yet the implication of the various arguments in this
volume is that in the 21st century, the Bank will not thrive
as an aid agency and that a continuation of business
as usual (a mix of functions and practices and habits
responding to multiple and varied demands, summarized
as “mission creep”) would deprive the global economy
of its continuing need for an institution dedicated to
shared prosperity.
The Bank is unlikely to achieve “relevance” in this more
global system, or “effectiveness” in helping countries
transform their economies, without the elusive “legitimacy”
it seems to have lost. The one step to furthering all three—
effectiveness, relevance and legitimacy—would be to ask
how as a club it might better serve all its members—rich
and poor.
Notes
1. Allan H. Meltzer, chairman, Report of the International
Financial Institutions Advisory Commission (Washington,
D.C., 2000). Also available at http://www.house.gov/jec/
imf/meltzer.pdf.
2. On the burden on developing country officials of
dealing with the resulting “cacophony” of donor voices,
views and demands, see Arnab Acharya, Ana Fuzzo de
Lima and Mick Moore, “The Proliferators: Transactions
Costs and the Value of Aid,” IDS Working Paper (Sussex,
U.K.: Institute of Development Studies, 2003).
3. The relevance of the non-borrowers’ guarantees
to the bank’s ability and cost of current borrowing rate
is sometimes questioned, since the bank’s financial
policies since the early 1970s have included substantial
provisioning and reserves, and since it has been so
rare that borrowers have delayed repayment let alone
CGD0502 0527_Engl_6x9.indd 82
8/17/06 2:55:11 PM
83
Selected Essays
defaulted. Since more than 80 percent of the IBRD’s
current reserves come from transfers from net income and
only a small minority from paid-in capital, there is a sense
in which the borrowers can now be said to be contributing
to the bank’s low borrowing cost and financial solidity.
Of course from the beginning the guarantees were, like
the nuclear option, only useful when not used.
4. Jessica Einhorn, “Reforming the World Bank,”
Foreign Affairs 84, no. 1 (2006): 18. The full quotation
is “Its core mission is no longer to partner with middleincome countries in their pursuit of balanced and externally
oriented growth; it is to alleviate poverty in poor countries
and in the poorest pockets of middle-income countries.”
I have omitted words to clarify that the key distinction
of interest is between partnership for externally oriented
growth and poverty alleviation—not only or necessarily
between middle-income and low-income countries.
5. Nancy Birdsall and Devesh Kapur, co-chairs, The
Hardest Job in the World. Five Crucial Tasks for the New
President of the World Bank, in this volume, p. 15.
6. See, for example, the World Bank Operations
Evaluation Department (OED), 2004 Annual Review of
Development Effectiveness (Washington, D.C.: World
Bank, 2004). The OED has now been renamed the
Independent Evaluation Group (IEG).
7. The debate about poverty or growth is often
decried as silly; for a good statement see Dani Rodrik,
“Growth Versus Poverty Reduction: A Hollow Debate,”
Finance & Development 37, no. 4 (2000). Birdsall
refers to “conditionality confusion” and suggests that
conditionality and ownership are complements, and not
in any way substitutes. See Nancy Birdsall, “The World
Bank of the Future: Victim, Villain, Global Credit Union?”
Carnegie Policy Brief No. 1 (Washington, D.C.: Carnegie
Endowment for International Peace, 2000).
8. See the Working Group Report in this volume for
more detail and for citations to these critiques.
9. Joseph Gold, Voting and Decisions in the
International Monetary Fund: An Essay on the Law and
Practice of the Fund (Washington, D.C.: International
Monetary Fund, 1972) p. 19. White’s statement is referred
to in this context by Ngaire Woods, The Globalizers: The
IMF, the World Bank and their Borrowers (Ithaca, New
York: Cornell University Press, 2006).
10. Gold 1972, p. 19.
CGD0502 0527_Engl_6x9.indd 83
8/17/06 2:55:11 PM
84
Rescuing the World Bank
11. Ahmed invokes Kemal Dervis, who provides
an extensive discussion and examples of how lack
of legitimacy is eroding effectiveness of the Bretton
Woods institutions. See Kemal Dervis with Ceren
Ozer, A Better Globalization: Legitimacy, Governance,
and Reform (Washington, D.C.: Center for Global
Development, 2005).
12. See also William Easterly, “What Did Structural
Adjustment Adjust? The Association of Policies and
Growth with Repeated IMF and World Bank Adjustment
Loans,” CGD Working Paper #11 (Washington, D.C.:
Center for Global Development, 2002).
13. See also Jeremy Bulow and Kenneth Rogoff,
“Grants versus Loans for Development Banks,” Paper
for AEA session on New Perspectives on Reputation and
Debt, January 7, 2005.
14. In addition to the Working Group Report, in this
volume, for a set of arguments see Nancy Birdsall, “The
World Bank of the Future: Victim, Villain, Global Credit
Union?” Carnegie Policy Brief No. 1 (Washington, D.C.:
Carnegie Endowment for International Peace, 2000);
and José Angel Gurria and Paul Volcker, The Role of
the Multilateral Development Banks in Emerging Market
Economies: Findings of the Commission on the Role of the
MDBs in Emerging Markets (Washington, D.C.: Carnegie
Endowment for International Peace, 2001).
15. A key rationale for the recommendation of the
Gurria-Volcker commission of differential loan pricing was
to encourage graduation of richer borrowers with access
to private markets.
16. The risk that China and other Asian nations will set
up a regional monetary fund is the key impetus currently
for the United States and Europe to agree to some change
in IMF quotas. A similar situation might arise at the Bank,
though in less dramatic form and creating less immediate
pressure on non-borrowers.
17. See, among others, Linn, who refers to the Bank’s
function as a “transmission belt” to carry low-income
countries through middle-income status into the ranks
of higher-income countries; Johannes. F. Linn, “The Role
of World Bank Lending in Middle Income Countries”
CGD0502 0527_Engl_6x9.indd 84
8/17/06 2:55:11 PM
85
Selected Essays
(comments presented at the OED Conference on the
Effectiveness of Policies and Reforms, Washington, D.C.,
October 4, 2004), p. 3; and de Ferranti, in this volume, who
refers to the Bank’s “broad-based analytical expertise on
development policy issues” and its ability to “combine
an appreciation of the broad macro perspective with
detailed examination of policy issues at the sectoral and
micro levels.”
CGD0502 0527_Engl_6x9.indd 85
8/17/06 2:55:11 PM
CGD0502 0527_Engl_6x9.indd 86
8/17/06 2:55:11 PM
Votes and Voice: Reforming
Governance at the World Bank
by Masood Ahmed
M
y objective here is to spark discussion about
the future governance of the World Bank group.
This is a large topic and one on which much
has been written. I am going to focus mainly on the role
and composition of the Board and its relationship with
management, but I recognize that there are also many
other aspects of the corporate governance agenda for
the IFIs.
Why Governance Matters
A legitimate initial question is whether improving the
governance of the World Bank is a priority issue for
delivering better development results for the world’s poor.
I start from the premise that the World Bank is the single
most important international actor in the development
business: ensuring its effectiveness over the coming
decade is a high priority for development policymakers
in rich and poor countries alike. And I am persuaded that
the World Bank cannot continue to deliver the results
we all want over the next decade without substantial
governance reform.
There are two broad sets of arguments that drive the
improved governance thesis. The first posits that voice,
legitimacy and effectiveness are mutually reinforcing
attributes for an international development organization,
not competing objectives.
Kemal Dervis, Administrator of UNDP, has argued
powerfully for the enduring merit of the U.N. in terms of
global legitimacy—including in a presentation to CGD.
He extended this line of argument to the Bretton Woods
institutions. In order for them to be a fully credible source
of advice, and for their advice to be backed by conditions
which would carry sufficient political acceptance to be
workable, he saw the need for a much greater sense of
Masood Ahmed is a national of Pakistan and holds an M.Sc. in Economics
from the London School of Economics. At the time of this symposium, he
was a Director General at the United Kingdom’s Department for International
Development. He has also held several senior positions at the World
Bank, and at the IMF where he currently serves as Director of the External
Relations Department.
CGD0502 0527_Engl_6x9.indd 87
87
8/17/06 2:55:11 PM
88
Rescuing the World Bank
their acceptance in the broader international community,
and particularly in developing countries. This acceptance,
he argues and I agree, is intimately bound up with their
legitimacy, in terms of how their governance is structured
and how that is perceived. So the argument is that it is not
good enough simply to have the right policy advice; that
advice is more likely to be accepted if it comes from an
institution that is seen as representative of the interests
of the borrowing countries.
These arguments apply to the World Bank’s work in
both middle-income countries, mainly through the IBRD,
and in low-income countries, mainly through IDA. A further
argument for improving developing country voice in IDA
stems from the ‘development coordinator’ role that it
plays for the broader donor community.
I now work for DFID. And right across the countries
where we operate, I am struck by the fact that there are
multiple donor interventions in the same area, overlapping
with each other, creating extraordinary demands on
scarce national administrators’ time and trying to get them
to focus on each donor’s strategic plan, each donor’s
set of conditionalities, each donor’s set of operating
specifications. Moreover, these countries frequently
don’t have the capacity to handle all these burdens
simultaneously. Middle-income countries usually do, but
many low-income countries, particularly in Africa, lack the
capacity to be able to provide a framework within which
each donor could operate in a highly complementary way
with the others and the national authorities.
Fortunately we are beginning to recognize the cost
imposed by this lack of harmonization and alignment.
Our pledge to do better is enshrined in the 2005 Paris
Declaration on aid alignment, harmonization and results.
However, this will be a long-haul endeavour, and for
many aid-dependant countries I see IDA as providing
essential backstopping to help governments to provide
a framework within which all donors can operate. Of
course, IDA will not be expected to do this alone—we
need to understand better how the U.N. system, the Bank,
the European Commission and a few well-positioned
bilaterals can be complementary—but it will be asked to
take on at least an important, highly visible and exposed
supporting role in very many country situations.
To play that coordinating role effectively, IDA needs
to enjoy external legitimacy, first and foremost with the
CGD0502 0527_Engl_6x9.indd 88
8/17/06 2:55:12 PM
89
Selected Essays
countries where it plays this role. It also needs to earn and
maintain legitimacy with other key development partners
and, significantly, opinion-shapers in donor countries
including in civil society.
A second important lens through which to consider
governance of the World Bank is from the perspective
of recent experience with corporate governance in the
for-profit and non-profit sectors. In the corporate world,
governance has evolved a long way in the last 20–30
years, as witnessed, for example, by the report of the
Cadbury Committee in the U.K. on this subject a few years
ago. While there is no cookbook recipe for governance of
a major corporation with many diffused shareholders and
millions of stakeholders, and no two situations are entirely
alike, there are some basic principles which command
widespread respect.
A key one is the importance of a relatively clean
delineation between the functions of management versus
non-executive directors (confusingly “executive” directors
in the Bank parlance). As Sir Adrian Cadbury pointed
out when he met informally with the Bank’s Board two
years ago, this basic requirement is not yet met in this
institution. More generally, it is not clear whether directors
are primarily operating in the narrow national shareholder
interest or for the wider corporate interest, and if the
latter, whether the process by which they are appointed,
retained and rotated favors or hinders this perspective.
Another obvious, and much analyzed, issue is the role
and selection of the Bank’s President, who combines,
U.S.-corporate style, the roles of Chairman and CEO,
an increasing anomaly on the other side of the Atlantic.
Therefore, both the changed international development
context and lessons from corporate governance argue
for change at the level of the Bank.
Building on Progress
The excellent recent report on the Bank by CGD identifies
some immediate improvements that are desirable and
feasible. Let me simply outline them before suggesting
two more fundamental ideas for change in the medium
term which need more exploration.
The first is to pursue disclosure more vigorously. The
Bank has come a long way in terms of disclosure, but
there’s still more that we can do to disclose country
strategies, and especially country level policy and
CGD0502 0527_Engl_6x9.indd 89
8/17/06 2:55:12 PM
90
Rescuing the World Bank
institutional assessment ratings. There are some valid
arguments for caution where the information may be
politically or market-sensitive, but the presumption must
be that these ratings can and should stand up to challenge,
and that sharing them increases the likelihood of positive
emulation. Over time, there is a case for progressively
externalizing the assessment function, using standards
developed by the Bank and other financiers against which
countries can then be benchmarked.
The second is the issue of decentralization. There
should be continued decentralization of decisions
as far as possible to country level, both in terms of
continued “deconcentration” of staff responsibilities
within the institution, but also by making more space
for effective decision-making by countries themselves.
The latter kind of decentralization requires everyone—
including shareholders—to accept a greater relativism
of policy options, to recognize that there are usually
several feasible adjustment paths from one situation to
another, and that weighing the pros and cons of each is
a sovereign matter. If the Bank, and for that matter the
Fund, come to be seen as more respectful of country
voice and ownership in this more fundamental sense, this
will improve governance and legitimacy in a major way
even without formal changes in the Washington-based
superstructure. This is significantly about changing the
day-to-day behavior of Bank staff, and so would require
a hard look at the Bank’s personnel management and
incentive framework.
The third area is trust funds, official-speak for
widespread ad hoc financing of the institution outside
of its core resources from capital and retained earnings
or, in the case of IDA, periodic core replenishments by
donor countries. I find it extraordinary that in a recent year
the World Bank received a larger sum of grants from its
shareholders in trust funds than it got for IDA. Some of this
money is for big multi-donor initiatives channeled through
the Bank (such as HIPC) or cofinancing for specific Bank
operations, but a substantial amount is for supplements
to the Bank’s own budget for policy or operational work.
It is remarkable that as shareholders we construct an
elaborate mechanism for setting priorities and discipline in
the Bank, and then as donors we bypass this mechanism
by setting up specific separate financial incentives to
try to get the Bank to do what we want. Inevitably this
CGD0502 0527_Engl_6x9.indd 90
8/17/06 2:55:12 PM
91
Selected Essays
is sometimes different from what the Bank’s Board, on
which we put our director, has just instructed it to do.
This is not unique to the World Bank by any means—in
the case of several U.N. agencies the accumulation of
such non-core or “project” funding has been larger overall
than core funding for years. And it is true that these
projects, on the whole, meet specific operational needs
and achieve results at their own level. But this constant,
sprawling, decentralized process of contracting for parallel
funding has a corrosive long term hollowing-out effect on
corporate governance. The pendulum has swung too far
in this direction and is overdue for a correction.
My fourth area for action is independent evaluation.
This is important in its own right as a key tool for improving
our understanding of what works in development, as we
discussed in another session in this conference. It also
has a huge payoff in terms of improving the legitimacy of
the policy prescriptions that come out of the institution
that is rigorously and publicly evaluated, in this case
the World Bank. It therefore enhances the process of
governance reform we have been discussing. I should
note that a focus on impact evaluation will also help
to raise the priority of improving the current woefully
poor quality baseline for development indicators in many
developing countries.
These are the types of immediate improvements we
should move forward on. But there are two more radical
questions which I’d like to explore.
The first question is whether it’s time to revisit this
model of a 200-person plus resident Board. It results in an
extraordinary degree of involvement of the shareholders
in the day-to-day management of the institution, wherein
the lines between management and shareholders begins
to blur.
Few corporations would consider having a permanent
resident Board of directors, let alone one like the Bank’s
which costs tens of millions of dollars a year to run, and
requires heavy dedicated management infrastructure to
service its requests for information. Even in the esoteric
world of public development finance institutions, especially
those created since the 1990s, this is a rarity. More often in
similar institutions, shareholders interact mainly through
brief periodic meetings of senior officials from capitals.
They can bring a more direct and authoritative connection
with domestic priorities, while complementary checks
CGD0502 0527_Engl_6x9.indd 91
8/17/06 2:55:12 PM
92
Rescuing the World Bank
and balances—such as robust oversight committees
on policy or top management appointments—operate
outside of the Board itself. I surmise that if the Bank
were re-created today, we would not invent anything like
the governance infrastructure that we have inherited.
It’s not obvious to me that moving over immediately
to a smaller, nonresident, non-executive board is the
only or best answer. There may be other solutions that
have equal merit. But the issue does need to be joined.
First, for cost reasons: taxpayers have a right to get
value for money. But also to get clear lines of corporate
governance responsibility.
The second question is: how do we introduce more
voices from developing countries in the different decisionmaking processes in and around the Board? Over the
past five years, there has been considerable work on this,
with proposals both covering the recalibration of relative
shares and voting rights and the suggestion to add one
or two African chairs to the Board to increase the voice
of the poorest countries. However, we are still short of a
consensus. Are we trying too hard for a one-size-fits-all,
comprehensive governance solution?
I’m beginning to come to the view that we have two
different problems of inadequate developing country voice
which need tailored solutions. One is that the emerging
markets don’t have adequate representation from their
perspective in IBRD, which is a kind of market-based
cooperative, of which they are an integral part. That’s a
different issue from the fact that the poorest countries,
particularly in Africa, who are the primary beneficiaries
and recipients of IDA financing, don’t have enough voice
in the IDA decision-making process.
The IDA problem is compounded by the role of IDA
Deputies, who increasingly set the framework within
which IDA operates, subject to later validation by the
Board with limited further debate. Although there are
now half a dozen borrower representatives as nonvoting
observers in the IDA Deputies’ meetings, they do not have
anywhere near the kind of intervention and capacity to
shape the policies of IDA. Donors have a right to insist
on value for money for their taxpayer investments in
IDA, but they should also get the best possible inputs
from recipient countries in determining what is and is
not effective, and this is not yet happening. Moreover,
when IDA management responds to Deputies’ (i.e., donor)
CGD0502 0527_Engl_6x9.indd 92
8/17/06 2:55:12 PM
93
Selected Essays
requests for policy papers, IDA recipients feel their ability
to shape the product is limited and ad hoc. Some have
also called for more of a challenge function, by asking
for ideas also from Southern development thinkers for
the Board and Deputies to consider.
If the formal Board voting structure were really the
essence of the problem, there are plenty of technical
solutions at hand. As Nancy Birdsall has pointed out, the
Inter-American Development Bank has a formula of full
parity in its voting between developed and developing
countries. Others—for example, IFAD with its three tiers of
capital representing recipient countries, OECD, and nonOECD donors—have found solutions that fit their political
needs. IDA itself has already provided for a potential
voting split which, while not quite 50-50, could go up to
48 (low and middle-income) to 52 (high-income). This
would involve poor countries taking up additional, heavily
discounted shares long reserved for them. The reason
they do not is revealing, and is arguably not primarily about
cost as this would be quite modest and could presumably
be subsidized further if needed. Rather it is because small
changes in the IDA voting shares alone would not affect
important decisions, such as constituency composition
which is driven by IBRD shares, or the relationship with
IDA Deputies.
So the question that we should be asking is whether this
diverse set of issues—including the under-representation
of the emerging markets in IBRD, and how IDA recipients
influence the shaping of IDA policies—can all be done
by trying to reconfigure shareholdings and/or adding a
couple of seats to the board of IBRD. I am increasingly
skeptical that a silver bullet exists.
Perhaps we need to step back and think more
fundamentally about whether we need a new framework
for the business of IDA, which brings in the donors to
IDA and the recipients of IDA, in a better-structured
conversation. They could shape rules and operating
criteria in which they would all feel more ownership.
This should be accompanied by a separate discussion
about the nature of emerging markets participation in
the market-based cooperative of IBRD. They need to
come together in a way that would represent ownership
from them of the role that IBRD plays in their economy,
and that they play in the world economy. This goes back
to the early history of IBRD as a cooperative tool for
CGD0502 0527_Engl_6x9.indd 93
8/17/06 2:55:13 PM
94
Rescuing the World Bank
the rebuilding of war-torn Europe, with its quite different
dynamic to that of an “aid agency.”
I freely admit I have not gone here beyond posing the
questions, and am offering no ready-made answers. I
also do not have a firm view yet on whether there is likely
to be enough political traction in the search for answers.
Moreover I fully appreciate how by differentiating between
parts of the Bank we could be raising thorny issues about
the relationship between these components, with the
potential loss of synergies embedded in the current setup. But I do think that this question of differentiating voice
in the IDA and IBRD contexts, along with the question
on the future of the non-resident Board, are fundamental
and we need to grapple with them.
I look forward to others joining this discussion in due
course, and thank CGD for giving the opportunity to
contribute these preliminary thoughts.
CGD0502 0527_Engl_6x9.indd 94
8/17/06 2:55:13 PM
The Battle for the Bank
by Ngaire Woods
T
he World Bank will not be able to avoid reform in
the immediate period. The powerful requestioning
of its sister institution in Washington DC cannot help
but spillover onto the Bank. Yet any “governance” reform
of the Bank needs undergirding with a clear sense of its
purpose and role. At least four kinds of Bank have been
skirted around. I suggest that we need a narrower test
to guide which of these Banks the World Bank should
become.
Reform is in the air
The winds around the Bretton Woods twins bode for
change. It is the IMF which is currently in the frontline
—not due to rabid criticism so much as a fatal lack of
interest. A few months ago I found myself a lone voice
among experts briefing the UK Treasury Select Committee
about globalization and the UK economy, arguing that the
IMF has an important role to play. The IMF has become
irrelevant opined the other experts. A few weeks later
speaking to the European Parliament, my case for
reforming the IMF fell on ears more receptive to the case
made by an Argentinian congressman—that the IMF had
destroyed Argentina’s prosperity. Inside the IMF there is
a scratching of heads. Is the public too ignorant or too
indifferent about the IMF? “We don’t know and we don’t
care” seems to be the public’s response.
Officials most closely involved with the IMF are pushing
to change what the organization does and how it is
governed. Most recently proposals have been made
by the United States, the United Kingdom, and South
Africa. US Treasury Official Tim Adams has argued for
Ngaire Woods, Oxford University and Center for Global Development
Ngaire Woods is director of the Global Economic Governance Programme
at University College, Oxford where she teaches and supervises research
in graduate international relations. She is an adviser to the United Nations
Development Programme’s Human Development Report Office, a member
of the Helsinki Process on global governance, and most recently served on
the three person external review panel asked by the IMF Board to review
the IMF’s Independent Evaluation Office. She sits on numerous editorial
and advisory boards, including the Advisory Group of the Center for Global
Development. Her most recent book is The Globalizers: the IMF, the World
Bank, and their Borrowers (Cornell University Press, March 2006).
CGD0502 0527_Engl_6x9.indd 95
95
8/17/06 2:55:13 PM
96
Rescuing the World Bank
weighted votes in the Fund to be altered to recognize
the growing economic strength of Asian countries in
the global economy, as well as for the IMF to have a
more ambitious and robust approach to exchange rate
surveillance.1 On governance, more radically, the UK’s
Central Bank Governor, Mervyn King, recently argued
that we should get rid of the Executive Board of the
Fund and instead use a non-resident Board meeting six
times per year.2 Meanwhile South Africa’s Central Bank
Governor has made the case for fundamental reform to
give developing countries more voice in decision-making
in the IMF, speaking—for all developing countries—of
the “highhandness,” “know-it-all approach” and “almost
patronising attitude towards developing countries” of the
institution.3
The World Bank will be affected
The arguments for reforming the IMF will affect the
World Bank on all three issues highlighted above. First,
the allocation of votes in the IMF has long been under
question but is now seriously under fire. The World Bank
will be directly affected by this debate since its voting
structure mirrors that of the IMF. Second, the governance
structure and the role of the IMF’s Executive Board is
rightly under question and this too will translate across
to the Bank which has the same basic permanent,
resident Board structure. Third, poorest borrowers have
long attacked the modus operandi of the IMF and its
ideological dogmatism. In its approach to lending, the
World Bank has taken serious steps to move away from
one-size-fits-all and to devolve ownership to its borrowing
members. But has it done enough?
The quest for change in the IMF will doubtless be
gingered by the fact that the institution’s largest borrowers
have been repaying, leaving the Fund short of prospective
income. But so too the Bank’s traditional borrowers are
enjoying access to alternative sources of finance and
turning away from the Bank—for reasons spelt out in the
CGD’s Report on the Future of the Bank. Both institutions
are having to explain afresh their raison d’etre. For the
Bank this entails asking—what makes us attractive? What
do we exist to do?
What should the Bank do?
In and around the World Bank a lot of effort has been
put into examining what the Bank does well and how it
CGD0502 0527_Engl_6x9.indd 96
8/17/06 2:55:13 PM
97
Selected Essays
might compete with other agencies. Yet the Bank is not
a private sector institution, nor a national or regional one.
It should not supplant the efforts of those competing
in the market. It is a public multilateral agency with a
fairly universal membership created by governments to
fill a need which neither private markets nor individual
governments can. What the Bank should be is what
it is uniquely placed to do as a universal, multilateral
development organization.
The Bank’s original mandate was to go where markets
were likely to fail to go—or fail to get to fast enough.
In 1944 this meant rebuilding Western Europe after the
second world war faster than nonmediated private capital
would. This would not only ensure economic growth
(and a market for US goods), but it would also alleviate
social and political fractures which would otherwise take
place. The Bank’s original mission also included lending
to developing countries whose immediate prospects
may not attract capital but whose stability and growth
were seen as important to global prosperity and growth.
Finally, the Bank was born to manage the excesses of
the market. This meant working to ensure an even growth
of trade so that the benefits of global commerce would
raise the standard of living and conditions of labor across
member countries. Put differently, the World Bank was
created (alongside the IMF) to manage what we now call
globalization, and in particular its “downsides.”
The original purposes of the Bank did not overlap or
contradict with what other international agencies would
be doing in the post-war period. Certainly the Bank was
born in a period of great belief about what governments
could achieve—after the New Deal and amidst the birth
of the welfare state in Britain. But in the intervening
years the Bank has become more of a jack-of-all-trades.
Paradoxically, the expansion of the Bank’s activities has
taken place alongside a proliferation of hundreds of other
multilateral agencies working on the same issues. For this
reason it is worth thinking harder about what the Bank
should and should not do—as indeed the CGD Report
on The Hardest Job in the World does. Let me propose
here a slightly more restrictive test than in that report,
aiming it at four of the kinds of Bank which were touched
on in that report.
CGD0502 0527_Engl_6x9.indd 97
8/17/06 2:55:13 PM
98
Rescuing the World Bank
The Knowledge Bank is the Bank which focuses on highquality research and its dissemination. Better collection
of data, research and sharing of information by the Bank
—we are told—will translate into better quality and more
even economic growth around the world. The same
rationale is used to justify multilateral surveillance and
research undertaken by the IMF. The assumption that
this role is necessary for each organization to achieve
its main goals is seldom tested. Of course knowledge
and furtherance of the social science of economics and
development economics is valuable, but it is not only
the Bank which is engaged in this. So too are the OECD,
universities around the world, regional and national policy
institutes, and other development organizations.
Should the World Bank be a “Knowledge Bank”? The
test is a two-fold one. First, is there something about
the way the Bank collects and disseminates knowledge
which is distinct from what other research and monitoring
organizations can do—and indeed which the Bank is
uniquely placed to do? Second, does the Bank’s work
as a Knowledge Bank contribute directly to its mandate
such as by contributing to more equitable and balanced
international trade so as to raise “the standard of living and
conditions of labor” across all of its member countries (to
quote Article 1 of the Bank’s Articles of Agreement)?
The Listening Bank is the World Bank of the Comprehensive
Development Framework and decentralization. It is a Bank
which listens more and imposes less. It puts borrowing
governments “in the driving seat” (in spite of the fact that
in none of those countries does the boss ever occupy the
driver’s seat). In some ways the “listening bank” tries to
reconcile “ownership” with the Bank’s ever-more intrusive
presence in borrowing countries. This could improve a
number of things about the Bank’s performance and
knowledge. It could enhance the Bank’s understanding
of how actual sectors in specific economies work—more
useful for the Bank’s poorest members than most of the
general theorizing done at headquarters. It could inject
some humility into Bank projects and policies—rendering
the Bank a genuine “development partner.” It could
also drag the Bank into all manner of projects including
democratization and social reform.
Should the World Bank be a “Listening Bank”? Again
the test is twofold. First, is the Bank uniquely placed
CGD0502 0527_Engl_6x9.indd 98
8/17/06 2:55:13 PM
99
Selected Essays
to engage in processes and policies implied in the
“listening bank” or is it supplanting what other organizations
(public and private) can and should be doing? Second,
will the Bank’s listening activities help it to fulfil its core
mandate? I am in no doubt that to some extent they will.
But the risk is that the trend to a greater presence on the
ground will tempt the Bank into an ever wider agenda,
full of good intentions but applying the wrong skills and
expertise to a mission which goes beyond that for which
it is equipped.
The Dams and Irrigation Projects Bank is the Bank
of those who want the Bank to attract back its largescale borrowers which give it a raison d’etre and a
healthy income stream. It is also a vision which pushes
back against the Bank’s widespread shift into more
easily-disbursed social and sectoral reform lending.
But dams, irrigation and large infrastructure projects
take the Bank squarely into a number of battlegrounds.
Procurement for large infrastructure contracts (including
by the world’s wealthiest countries and corporations)
is notoriously rife with corruption and kickbacks. It sits
with difficulty alongside the anti-corruption goals of the
new President. Large infrastructure projects often strip
people of their homes and damage local environments,
leading the Bank into head-on conflicts with NGOs and
local communities which it has been (and still is) trying
so hard to cultivate.
Should the Bank be engaged in large-scale infrastructure
lending? Tough as it may be, this was a key part of the
original mandate of the Bank which includes developing
productive capacity in countries when private capital is
absent or too expensive. The uniqueness of the Bank’s
role here stems from its capacity to raise finance more
easily and cheaply than any individual country. But do
its activities in this area contribute to fulfilling the Bank’s
mandate? Recall that the Bank’s mandate is not simply
to promote economic growth but to promote trade,
investment, and productivity which is balanced and
contributes to better standards of living and conditions
of labor within and across its members. The Bank has
a duty—distinct from the private sector—to ensure that
all people’s living standards and working conditions are
bettered. At the very least it should work in ways which
CGD0502 0527_Engl_6x9.indd 99
8/17/06 2:55:14 PM
100
Rescuing the World Bank
ensure that basic human rights are not impinged. This is
difficult but has to be part of what the Bank does.
The Big Expensive Bank. Finally, it is worth mentioning
the Big Expensive Bank which has little by way of a hard
budget constraint (for increases in costs can be passed
on to borrowers). It is a Bank which has spent millions
on advice and restructuring within its own walls as each
new President has attempted to recreate around himself
an institution with which he is more familiar. There is no
rationale for an undisciplined Bank. The Executive Board
and Board of Governors (member countries) have robust
powers of oversight which they can exercise. The problem
is that they too seldom, and too ineffectively do so.
Governance reform is inevitable
The Boards of the Bank (both the Board of Governors
and the Executive Board) need to act properly as
supervisers of the institution rather than micromanagers.
Their job is to ensure that the Bank fulfils its strategic
goals—defined here in terms of what the Bank is uniquely
placed to do. Other contributors have commented on
governance reform in the Bank, and I have written
extensively about it elsewhere. Suffice to say here that
current arrangements have proven to be ineffective
from a corporate governance point of view as well as
from a political “legitimacy” point of view. The senior
management is selected by a process seen as neither
fair nor meritocratic by the rest of the world and results
in a Bank unduly skewed towards its largest vote-holder.
Although small and expert, the Board has not been an
effective strategic arm or constraint on the management
of the Bank. Nor is the Board seen adequately to represent
the full membership of the Bank whether seen in terms
of economic weight, affectedness by Bank actions, or
contributors to the Bank’s expenses.
There are some straightforward solutions to these
problems.4 First, on leadership the President and senior
management must be seen as equally accountable to all
countries who are members of the Bank. All countries
pay for the institutions—they should also all have a say.
The Bank sits in Washington DC and therefore is prima
facie perceived as primarily accountable to the United
States. Its President and Senior Management need
powerfully to balance that perception. At the very least
the double-role of the US as host of the institution and
CGD0502 0527_Engl_6x9.indd 100
8/17/06 2:55:14 PM
101
Selected Essays
holder of the Presidency needs reducing by dropping
the convention that the US appoints the President or
by shifting the Bank’s headquarters to another capital
(which would be required by the Articles if the Europeans
sat with one seat on the Board and thereby became the
largest quota holders).
The Board needs to be effective in overseeing and
monitoring management, and ensuring that the Bank’s
core activities adapt appropriately to reflect what the
Bank is uniquely placed to do and what contributes
directly to its mandate. At present the Board neither
represents the Bank’s membership adequately nor fulfils
these core functions. The Board has eight Directors
which directly represent individual countries (United
States, France, United Kingdom, Japan, Germany,
Russia, China, Saudi Arabia), and sixteen Directors
who represent the rest. Most Directors live in a grey
zone, based in Washington DC, paid by the Bank, and
neither instructed by, nor accountable to, most of the
membership of the Bank.5
For about 174 members of the Bank, there is little
incentive to engage in decisions being made by the
Board. This is because eight Directors can marshal a
majority among themselves with little if any consultation
with others. This does not have to be the case. If
Directors had to marshal not just 50% of votes (which
might be just 8 members), but also 50% of members
(92 countries) to make decisions, there would be a
clear incentive to consult and bring on board Directors
who represent a large number of countries but wield
few votes (such as the two Directors who represent
over twenty African countries each yet each wield
less than 3.5% of voting power). This is not a difficult
reform. The Bank’s Articles already provide for doublemajority voting (Article VIII) for any amendment to the
Articles. This could be extended to other decisions.
Along with transparency of the Board’s process such
as publication of the full minutes of any Board meeting
so that countries can read exactly what their Director
has said in Board meetings, these would be first steps
towards a more effective Board.
CGD0502 0527_Engl_6x9.indd 101
8/17/06 2:55:14 PM
102
Rescuing the World Bank
In brief, reform is in the air around the IMF and World
Bank in Washington DC—and so it should be. Powerful
members of the Bank should be pushing a new more
effective structure of governance, and a strengthening
of the unique contribution the World Bank can make
to the equitable spread of globalization and economic
growth.
Notes
1. Timothy D. Adams (US Department of the Treasury),
The US View on IMF Reform, Speech presented at the
Conference on IMF Reform, Institute for International
Economics (Washington, D.C., 23 September 2005).
2. Mervyn King (Governor of the Bank of England),
Reform of the International Monetary Fund, Speech
given at the Indian Council for Research on International
Economic Relations (New Delhi, 20 February 2006).
3. Tito Mboweni (Governor of the Reserve Bank of
South Africa), Speech at University of Pretoria (Pretoria,
28 February 2006).
4. These are elaborated at greater length in chapter
7 of Ngaire Woods, The Globalizers: the IMF, the World
Bank, and their Borrowers (Ithaca, New York: Cornell
University Press, 2006).
5. For a closer analysis of the functioning of the
Boards see Ngaire Woods, “Making the IMF and World
Bank more accountable,” International Affairs 77,
no. 1, (January 2001): 83–100; and Ngaire Woods and
Domenico Lombardi, Uneven patterns of governance:
how developing countries are represented in the IMF,
Review of International Political Economy 13, no. 3
(August 2006): 480–513.
CGD0502 0527_Engl_6x9.indd 102
8/17/06 2:55:14 PM
The World Bank and Low-Income
Countries: The Escalating Agenda
by William Easterly
I
have a very simple message about the World Bank
and low-income countries. To be effective, the World
Bank needs to have in place a set of tasks, a mission,
and an incentive system that will create accountability for
results. Accountability for results implies that there will
be some reward for getting results and some penalty for
not getting results. That’s the first message.
The second message is dispiriting. The World Bank in
low-income countries is now and has been for a long time
suffering from a really bad case of mission creep. Such
mission creep has taken it farther and farther away from
tasks on which it is even feasible to have accountability.
To reverse that trend is really the first step in having a
World Bank that is accountable for achieving results in
low-income countries.
Let me tell you what I mean by “mission creep” and
how this is a long-run tendency. There have been a lot of
interventions that have been tried, and these interventions
have been unsatisfactory, leading the World Bank to try
an ever more ambitious and extensive set of interventions
in an attempt to make up for the failure of the previous
interventions. To get concrete, in the early days, The
World Bank was all about roads, dams and schools. The
mentality was, if you build it, they will come—by doing
specific projects that would create tangible outputs.
To be fair, there was some success in that period. There
were successes at building roads, dams and schools.
There were successes at building infrastructure for clean
William Easterly, New York University and
the Center for Global Development
William Easterly is Professor of Economics (joint with Africa House) at New York
University and a non-resident fellow at the Center for Global Development.
He spent 16 years as a research economist at the World Bank. He has worked
in many areas of the developing world, most extensively in Africa, Latin
America, and Russia. His areas of expertise are the determinants of long-run
economic growth and the effectiveness of development assistance efforts.
He is the author of The Elusive Quest for Growth: Economists’ Adventures
and Misadventures in the Tropics (MIT, 2001), The White Man’s Burden: Why
the West’s Efforts to Aid the Rest Have Done So Much Ill and So Little Good
(Penguin, 2006) and numerous articles in leading economics journals and
general interest publications.
CGD0502 0527_Engl_6x9.indd 103
103
8/17/06 2:55:14 PM
104
Rescuing the World Bank
water, and indicators of access to those services have
actually gone up in Africa. In fact, one of the success
stories that is not talked about enough in the literature is
that there have been some major achievements in Africa
and they usually have to do with the specific project tasks
that used to be the main mission of the World Bank.
There was also a certain level of dissatisfaction that
such specific projects did not bring rapid development to
Africa, and so then we move to the next phase. The idea
took hold that to be effective, project interventions needed
to take place in the presence of good macroeconomic and
microeconomic policies, mainly free markets, free trade,
low government deficits, macro stability—the Washington
consensus. This line of thinking brought us to the age of
structural adjustment.
The World Bank, beginning in 1980, started making
loans conditional on adopting a large number of policy
reforms in an attempt to redress what was perceived
as a gap in the previous effort, or as a reason why the
previous effort failed: that the interventions were not
themselves enough to create development, because if
the overall policy environment was so badly distorted,
then development would not happen.
Unfortunately, structural adjustment did not work.
Growth did not happen. Policy reforms were very erratic
and uneven. The consensus in the scholarly literature
currently is that, overall, structural adjustment lending
failed to attain its objectives.
If you want to go beyond the scholarly literature, then
there is a simple stylized issue of real import, which is
the main topic on the agenda of the IMF-World Bank
meetings this weekend, and that is debt relief.
Debt relief is talked about as sort of this benevolent
thing that rich countries are doing for Africa, but what
we forget is the flip side of debt relief, that debt relief is
really a sign of the failure of structural adjustment.
It is precisely those countries that got a lot of structural
adjustment loans that became HIPC (Heavily Indebted
Poor Countries), and thus found themselves in need of
debt relief. The fact that they could not pay back zero
interest loans with a 40-year maturity is itself a completely
compelling sign that structural adjustment lending had
failed in these countries.
That is true not only of the obvious failures that had
negative growth in Africa (which are the majority of those
CGD0502 0527_Engl_6x9.indd 104
8/17/06 2:55:15 PM
105
Selected Essays
who received structural adjustment loans). It is also true
of those that are touted as success stories of structural
adjustment—countries the likes of Uganda and Ghana
are also getting debt relief. Even the success stories have
not been able to pay back structural adjustment loans.
Then we have the next wave of escalation, which
maintained that if policies were not the answer, then
institutions must be, meaning that the task of international
agencies like the World Bank should be to try to promote
change and progress in institutions.
There are also various kinds of new vehicles loosely
related to promoting institutions, like the new poverty
reduction strategy papers, and, in general, all kinds of
initiatives to try to promote institutions. I have difficulty
understanding exactly how these initiatives are supposed
to work, for the simple reason that no one really knows
how to change institutions from the outside.
We can think of specific piecemeal things that would
change institutions for the better. However, we really do
not yet know how to achieve a wholesale transformation
of institutions. The evidence that we have is that, if
anything, aid leads to a worsening of good governance
in recipient countries.1
That is essentially where we are now, except in some
countries there has even been a further step, indeed
an even more ambitious one, which is in the states that
are called “failed states.” It is in these countries where
we have even more escalation, where international
organizations would actually take over some of the
functions of government and have some kind of new
trusteeship type arrangement in failed states. This is
exceedingly difficult, no doubt, and yet a further escalation
of the current thinking that if we can’t change institutions
from the outside, let’s ourselves become the institutions
and take over.
What, then, is going on here? Why do I say that
escalation has taken us farther, and taken the World Bank
ever farther and farther away from being accountable?
Well, there are two big problems with this escalation.
One is measuring results and the other is what the results
depend on. Measuring the World Bank’s effort and second,
what the outcomes depend on. So you can see with each
successive step, it becomes more and more difficult to
measure the World Bank’s effort.
CGD0502 0527_Engl_6x9.indd 105
8/17/06 2:55:15 PM
106
Rescuing the World Bank
It was easy to measure the World Bank’s effort when
it was just building a road or building a dam—something
highly visible, that could be monitored and accounted
for—and it would be very embarrassing if the World Bank
gave money for a road and it was not built.
With structural adjustment it became much more
difficult to hold the World Bank accountable for the
goal of changing policies and the effort that goes into
changing policies.
Of course, the natural question is: how do you
measure what policy would have been without World
Bank intervention? How do you actually measure changes
in policies? Policies are a very ill-defined set of lots of
different actions.
With institutions, it becomes even more difficult to
measure progress, and of course with post-conflict
reconstruction, where you’re trying to change everything,
then the game is up and it becomes hopeless to try to
measure the World Bank’s contribution.
The second problem for accountability with the mission
creep of the World Bank is that with each step of escalation
the results depend on more and more factors besides
simply what the World Bank does. When the World Bank
is building a road, the outcome mainly depends on the
actions of the World Bank itself. The World Bank can
pretty much determine whether a road gets built or does
not get built.
Yet with changes in policies, there are now many
actors trying to influence the country to change even
more policies, and more so with institutions. Such
changes depend not only on outside factors but also
domestic political factors, meaning that it becomes
increasingly routine to escape accountability because if
something goes wrong, everybody can point fingers at
everybody else.
The World Bank can say it is the IMF’s fault, the IMF
can say it’s the World Bank’s fault, or they both can say
it was the recipient government’s fault. Alternatively,
the recipient government can just say it was the fault of
politics that left them powerless.
Blame can also be leveled against outside factors like
a hurricane or terms of trade shock, meaning that nobody
can be held accountable for the results in these areas. So
it becomes overly burdensome to simply hold the World
Bank accountable for efforts in these areas.
CGD0502 0527_Engl_6x9.indd 106
8/17/06 2:55:15 PM
107
Selected Essays
To make matters worse, with each step it becomes
more difficult to prove that the World Bank can even
affect these outcomes. With infrastructure projects, for
example, it is fairly obvious that the World Bank can
build a road, if it wants to. There may be problems with
maintenance, but those problems could be solved with
particular measurable efforts.
The evidence on changing policies is much weaker.
Pretty much the result is, aid does not change policies.
The evidence is also such that aid does not change
institutions, at least not for the better. It may change them
for the worse, and God only knows what we may find ten
years from now of the effect on post-conflict countries,
of whatever the effort of the World Bank was.
With all of these problems, it seems we have moved
ever farther and farther away from accountability. What
needs to happen to reverse this trend of escalation and
why has escalation happened?
Let’s go back to the scholarly literature on development.
One thing emerging more and more from the literature
is that aid, and outside actors like the World Bank,
cannot achieve development and transformation of other
societies like poor countries in Africa. This is still a very
controversial conclusion but seems to be borne out by
a lot of evidence, such as the poor track record of the
escalating interventions.
Development and transformation is just something
that outside actors cannot achieve. They do not have
the tools, the ability, the incentives, and accountability
in place to do so effectively. Even if they did have the
incentives and accountability, it is not clear that outsiders
like the World Bank can achieve wholesale transformation
of another society like the complex societies in lowincome Africa.
Does that mean that all is lost? No. There are still a lot
of good things that aid can do. Aid can do those small
piecemeal things like building roads and maintaining
them, like getting 12-cent medicines to children who
would otherwise die from malaria, like sinking boreholes
and getting clean water so that people don’t get sick
from contaminated water.
Aid can do small-scale things like the current project in
the World Bank on the costs of doing business, in which
you try to take specific piecemeal steps to lower the red
tape of doing business in developing countries. Doing
CGD0502 0527_Engl_6x9.indd 107
8/17/06 2:55:15 PM
108
Rescuing the World Bank
this creates opportunities for what is the real engine of
development and transformation in low-income countries,
which is the private sector and it’s the homegrown efforts
of local people themselves, both political leaders and
private sector within African countries, within low-income
countries themselves.
That is where development and transformation is going
to come from, not from the outside actors.
If the World Bank is willing to focus on those more
modest tasks, then feedback and accountability is feasible
and the World Bank could achieve better results. If the aid
community focused on these smaller tasks and held the
World Bank accountable for achieving those results the
roads that are not being maintained now actually could be
maintained. The 12-cent medicines that are not reaching
the children dying of malaria could reach the children
dying of malaria, if the aid community focused on simple
tasks like that and held the World Bank accountable for
those tasks.
Otherwise, development will happen mainly because
of what Africans do, because of what people in lowincome countries do, and not because of what the World
Bank does.
Notes
1. See Stephen Knack, “Aid Dependence and the
Quality of Governance: A Cross-Country Empirical
Analysis,” Policy Research Working Paper 2396,
(Washington, D.C.: World Bank, 2000).
CGD0502 0527_Engl_6x9.indd 108
8/17/06 2:55:15 PM
The Role of the Bank in
Low-Income Countries
by Steven Radelet
T
his note makes four brief points about the role of
the World Bank in low-income countries. The first
point concerns mission creep, or lack of institutional
discipline. The Bank is involved in too many activities in
individual countries and does not have a particularly clear
focus. This lack of focus and overextension transfers to
the recipient country governments who are encouraged
by the Bank and other donors to take on way too many
issues and activities, leading to a lack of focus, no sense
of priorities, and less progress on a small number of really
critical issues.
The Bank does this partly because it has, in house, a
wide range of expertise and a decentralized structure so
that it tends to try to support all kinds of activities. The
main concern is not necessarily that the Bank globally has
expertise in too many areas and needs to narrow its focus
as an entire organization, although that is an issue. The
more important problem is that within individual countries
it has difficulty focusing on the really key issues, deciding
that some problems cannot be solved right away, and
determining a small number of very high priorities. It needs
to do a much better job of both setting its own priorities
within countries and helping recipient countries think
through their priorities.
As a result, the Bank and other donors also tend to
encourage an attitude of trying to solve all problems at
once. It is very easy to go into a low-income country and
find 25 or 50 or 60 problems, and to tell the recipient
country that X is not working very well, we’ve got to fix
Y, we’ve got to change Z, and so on. While it is true that
these may all be problems, there is no sense of priorities;
the problem is that in developing countries, the scarce
Steven Radelet, Center for Global Development
Steven Radelet is a senior fellow at the Center for Global Development,
where he works on issues related to foreign aid, developing country debt,
economic growth, and trade between rich and poor countries. He is coauthor of Economics of Development (a leading undergraduate textbook),
and author of Challenging Foreign Aid: A Policymaker’s Guide to the
Millennium Challenge Account (Center for Global Development, 2003).
CGD0502 0527_Engl_6x9.indd 109
109
8/17/06 2:55:15 PM
110
Rescuing the World Bank
resource is strong government policymakers, meaning
that there are only so many things that can be tackled.
The real challenge in development policy is not finding
problems, but determining which problems should be
solved first, given limited resources, to get the biggest
bang and set the stage for continued change.
The good news is that to achieve development, we
do not have to solve everything at once. If you look at
the countries that have been successful over the last
40 years, mostly but not exclusively in Asia, they have not
solved everything at once.1 Take China as an example.
No one would argue that China has solved everything at
once. Nor have Korea, Indonesia, Malaysia, Botswana,
Mauritius or Chile.
These and other successful countries were able to
set priorities and get a few critical things right, and really
solve some of the most pressing problems, rather than
attempt to solve a wide array of problems simultaneously.
Unfortunately, the Bank and the donors do not do this very
well. Most specifically, the World Bank’s Comprehensive
Development Framework is a mistaken approach, because
it encourages the attitude of “let’s try to solve everything at
once and fix all of these problems because development
is so complicated,” rather than “let’s set some priorities
and try to make real progress on the most important
issues first, and follow on with others later.”
Second, the Bank and other donors have to do a better
job of recognizing that not all developing countries are
alike, and it is necessary to differentiate the strategies
that are used within developing countries. Here I am
not making the point that the substantive development
priorities differ across countries, which they obviously do,
but rather that the quality of governance, commitment
to development, and institutional capacity differ widely
across countries, so the approach the Bank takes should
recognize these differences, and the Bank should more
clearly vary the way it provides its assistance across the
spectrum of countries. There has been a lot of talk in the
last few years about selectivity, the principle being that
we give more aid to countries with better governance
and institutions and less to countries that don’t perform
as well. This is a sensible starting point, but we need
to go beyond that and actually deliver aid differently to
countries that have different capabilities and qualities
of governance.2
CGD0502 0527_Engl_6x9.indd 110
8/17/06 2:55:16 PM
111
Selected Essays
For example, in recent years there has been a lot of talk
surrounding the issue of budget support, about country
ownership, about longer-term commitments, about all
kinds of things, as if these are the right solutions for all
developing countries. But they are not. The principle of
country ownership may not be appropriate under certain
circumstances. Zimbabwe, with its current destructive
government, is a perfect case in point—no donor should
give more ownership of the development program to the
government of Zimbabwe, and none really does. But our
rhetoric about improving aid effectiveness does not take
these differences into account, and implies that changes
that make sense in some countries are a good idea for all.
Donors, including the World Bank, need to move beyond
the general rhetoric and shift toward thinking about how
to employ different kinds of approaches and modalities in
different countries. In countries that have better governance,
better institutions, and have shown some commitment
toward progress—countries such as Ghana, Honduras,
Mongolia, Mozambique, Senegal, and Tanzania—it makes
sense to have more country ownership, to provide longerterm commitments, to send more of the money through
the budget as budget support, and otherwise change the
ways that we deliver assistance.
In countries with weaker governance, we should stick
with more project support, look for a narrower set of
activities, and have a mix of donor priorities and country
priorities. In countries with the weakest governance,
there should be a much narrower set of activities, much
less government ownership and involvement in setting
priorities, a shorter time frame, shorter time commitments,
much tighter oversight in what is done, a different way to
measure results and different ways of delivering money,
with less of it through the government and more of it
through non-governmental organizations.
We have to shift toward creating these more distinctive
strategies for different countries. Donors are beginning
to move a little bit in this way, in some cases implicitly
and in others more explicitly. The United Kingdom’s
DFID and some other European donors are providing
budget support in some countries but not in all (although
the criteria they use to make these distinctions are not
clear). The United States has set up the Millennium
Challenge Account, which very explicitly distinguishes
among recipient countries. And there is some welcome
CGD0502 0527_Engl_6x9.indd 111
8/17/06 2:55:16 PM
112
Rescuing the World Bank
movement within the Bank along these lines, but it needs
to go further in providing its assistance in different ways
across countries. The Bank’s increased use of grants
opens many new possibilities that it has not yet begun
to explore about to whom and how it provides financing
under different circumstances.
Establishing more distinct modalities could help create
incentives for countries to strengthen governance and
institutions. Budget support provides a good example.
It makes sense for the Bank to provide budget support
in countries that have better public sector finance
institutions, stronger fiduciary standards, and better
accounting and auditing practices, but not in countries
with weaker systems (note that this is not the same as
better governance more broadly). There are now many
ways of ranking and grading public sector finance
institutions, such as the budget office and the ministry
of finance more generally. The Bank should use such
grading systems, and provide a greater share of its
funding as budget support to countries that score better
on these standards. For example, as governments reach a
certain standard on auditing, accounting, publishing their
accounts, procurement, and other areas, they receive
some share of their funding as budget support, say
20 percent. As their standards improve, they could receive
50 percent or 75 percent or more. Note the issue is not
about how much money they would receive, but how
they would receive it. The incentive would be built in for
the countries to want to improve their systems, because
by doing so they could receive more of the money in the
way they want it—as budget support.
The third point is on accountability. The Bank currently
does not reward results strongly enough, and too often
it rewards failure. It needs to be much more resultsoriented. To its credit, the Bank has begun to move in
this direction in recent years. But it is a huge challenge
to try to change incentives within an institution, and to try
to reward success rather than failure. Part of the answer
is in removing long-standing incentives for Bank staff
that are focused on disbursing money, and creating
incentives that are connected to the success or failure
of the activity. But these changes will not just happen by
people saying we ought to do a better job and we ought
to focus more on results and by hoping staff respond.
Instead, changes must come from the Executive Board
CGD0502 0527_Engl_6x9.indd 112
8/17/06 2:55:16 PM
113
Selected Essays
and from senior management who must make structural
and policy changes that create incentives that are focused
on results. Management could make proposals to the
Board for re-orienting staff incentives and promotion over
time so that they are more linked to results, or to how
projects are monitored and evaluated over time to focus
on results. Similarly, Board members—both contributors
and recipients—could demand more results-based
approaches.
The United States has been pushing from the Executive
Board to hook its IDA contributions to broad indicators
of the Bank achieving results. I am generally supportive
of this approach, although I have not agreed with all the
details of how it has been carried out. It would be better if
these kinds of initiatives primarily came from management
rather than the Board. It will also be a challenge to translate
these kinds of measures from an institution-wide focus
to specific projects and programs.
President Wolfowitz has made the point about the need
for the Bank to be more results-oriented, and hopefully
he will be able to move the Bank in this direction. One
key in focusing more on results, as proposed by my
colleague Nancy Birdsall, would be the creation of a
truly independent outside evaluation entity that can
measure results on specific activities and for the Bank
more generally.3
The fourth and last point concerns grants. From the
Bank’s perspective, providing more of its assistance to
low-income countries as grants makes a lot of sense, but
the way the Bank is allocating grants across countries
doe not make much sense.4
When the shift to grants began in 2002, the Bank
decided to allocate grants based on sector—certain
activities, such as health education received more grants,
while others such as infrastructure were financed by
loans. It quickly became apparent that this formulation
would not work, as countries would receive mixes of
grants and loans, leading to incentives to move money
and creating confusion.
More recently, the allocation rule was changed so that
the share of grants a country receives is based on measures
of debt sustainability. The rationale seems simple: grants
were pushed by the United States and others partly as a
solution to sustained debt problems, and giving grants to
countries with the largest debt problems certainly helps
CGD0502 0527_Engl_6x9.indd 113
8/17/06 2:55:16 PM
114
Rescuing the World Bank
reduce their future debt burden. But basing grants on
debt sets up strong perverse incentives, because the
more debt a country accumulates, the more likely it will be
to receive grants, whereas the more a country manages
its economy well and avoids debt problems, the more it
will be told that it must continue to borrow. This rationale
does not make a lot of sense, especially as we are about
to forgive all the debts for countries that reach the HIPC
Completion Point, meaning that those countries will be
prime candidates for more loans, rather than grants.
Instead, grants should be based on a country’s income.
The poorest countries in the world should get grants,
and as their incomes grow, they should receive more
loans—first subsidized, and later not subsidized. That’s
the principle the Bank uses to distinguish between IDA and
IBRD funds, and that’s the way most donor flows work.
As incomes grow and countries achieve higher incomes
that demonstrably prove a greater ability to service debt,
the level of concessionality should decline.
The poorest countries in the world are the ones that
face the deepest development problems, so they face
the greatest risk that they will not be able to achieve
the growth necessary to repay loans, even if they
establish good policies. They tend to be vulnerable
to the greatest number of shocks and face the largest
obstacles to growth. When they do achieve growth, the
resources should be reinvested, not repaid to the Bank
to be relent elsewhere.
Going forward, the Bank should set up either a third
separate window or a window within IDA for grants only
for all countries with incomes below $500 per capita. This
would ensure that the poorest countries receive the most
concessional money and do not face debt problems. As
incomes grow in these countries and they begin to show
some capacity to actually get returns on investments,
they can go to IDA loan-financing. This approach will
better match grant financing with the greatest needs,
and avoid the perverse incentives of allocating grants
based on debt.
CGD0502 0527_Engl_6x9.indd 114
8/17/06 2:55:16 PM
115
Selected Essays
Notes
1. See Steven Radelet and Jeffrey Sachs, “Reemerging
Asia,” Foreign Affairs 76, no. 6 (November/December
1997): 44–59.
2. Steven Radelet, “From Pushing Reforms to Pulling
Reforms: the Role of Challenge Programs in Foreign
Aid Policy,” in The New Public Finance: Responding to
Global Challenges, eds. Inge Kaul and Pedro Conceição,
(Oxford: Oxford University Press, 2006). Also available
as CGD Working Paper #53 at http://www.cgdev.org/
Publications/index.cfm?PubID=196.
3. Nancy Birdsall, “Seven Deadly Sins: Reflections
on Donor Failings,” CGD Working Paper Number 50,
December 2004.
4. See Steven Radelet, “Grants for the World’s Poorest:
How the World Bank Should Distribute Its Funds,”
CGD Note, June 2005, http://www.cgdev.org/content/
publications/detail/2681.
CGD0502 0527_Engl_6x9.indd 115
8/17/06 2:55:17 PM
CGD0502 0527_Engl_6x9.indd 116
8/17/06 2:55:17 PM
Has the World Bank Lost Control?
by Adam Lerrick
“We are facing…competition [from the capital
markets]. I think it’s important that we effectively
compete. Increasingly,…if the fight against poverty
is successful, more and more countries will be
in this middle-income category, and if this
institution is going to remain relevant to the world,
it obviously needs to be relevant to the middleincome countries.”
World Bank President Paul Wolfowitz, September 22, 2005
T
he World Bank is in big trouble. Major middleincome countries, the cream of the Bank’s portfolio,
are curbing their borrowing and paying down their
balances, setting off alarms at the Bank. Net loan flows
have shifted $30 billion over the last seven years, from
positive to negative. Instead of drawing a net $14 billion
from the Bank in 1999–2002, these nations repaid a
total of $15 billion in 2003–2005. The cause is clear: The
interest subsidy embedded in Bank loans, a compelling
12 percent per annum on average in 1999, has now
shrunk to less than 2 percent as emerging nations have
gained increasingly greater access to private capital.
The difference is no longer enough to persuade finance
ministers to realign their economic priorities with the
social agendas of the Bank’s rich members.
For years, the Bank has been in the business of lending
at highly subsidized rates to non-needy nations. Ninety
percent of Bank loans now go to just 27 borrowers, 10
of these accounting for 75 percent, a list that closely
parallels private sector choices, and for these nations
the Bank contributed a mere 1 percent of the average
Adam Lerrick, Carnegie Mellon University
and American Enterprise Institute
Adam Lerrick is the Friends of Allan H. Meltzer Professor of Economics at
Carnegie Mellon University and director of the University’s Gailliot Center
for Public Policy. He is also a Visiting Scholar at the American Enterprise
Institute. His published papers include “A Leap of Faith for Sovereign
Default: From IMF Judgment Calls to Automatic Incentives”, “The World
Bank as Foundation: Development without Debt” and “Bank Deposit
Receipts: A Transitional Money for Financial Crisis”. Lerrick is an expert in
monetary economics, international finance, financial markets and the role of
international financial institutions in the world economy.
CGD0502 0527_Engl_6x9.indd 117
117
8/17/06 2:55:17 PM
118
Rescuing the World Bank
net $200 billion that the capital markets have provided
each year over the last decade.
When the International Bank for Reconstruction and
Development (IBRD) was founded and its self-image
formed, capital markets were small, segmented and
cautious. The Bank was to borrow in the markets,
backed by the AAA guarantee of its rich industrialized
membership, and lend on to developing countries that
could not access resources to fund growth. International
financial intermediaries to channel funds and assume
risk were in short supply. The plan was for developing
economies to be nourished only until they had gained
the financial credentials to attract private capital on their
own. This was called “graduating.” But the Bank won’t
let go.
The Bank was enjoined from displacing the private
sector. Now it wants to compete. With its monopoly power
lost, the Bank is scrambling to maintain market share by
lowering prices. In the end, it is the demands that are
at the very center of its mission that will be sacrificed to
maintain competitiveness.
Middle-income borrowers are clearly good for the Bank.
The Bank wants to keep its best, lowest risk customers.
Their loans are more likely to be paid and their projects
more likely to succeed. Without these prime clients to
raise the value of its portfolio, both its credit and its
credibility would be challenged. And the Bank has been
willing to pay for the privilege of “staying involved.” Over
the past twelve years, IBRD loan revenues have fallen
short of administrative costs by a cumulative $3 billion.
Over time, more and more countries will move into the
middle-income group that already commands two-thirds
of World Bank Group money and effort.
But is the World Bank good for middle-income
borrowers? The Bank’s litany of reasons for lending is
refuted by the facts of the market place. Its premise of a
shortage of private funding is no longer valid. Its business
model that relies on subsidized financing is outdated. Its
advice now has a negative value to its best clients. In
a world of finite aid resources, its money and effort are
better dedicated to the poorest nations whose access
to private capital is in the distant future.
The global economy has out-distanced the need
for World Bank lending to emerging nations and along
the way has done more to alleviate poverty than all the
CGD0502 0527_Engl_6x9.indd 118
8/17/06 2:55:17 PM
119
Selected Essays
interventions of officialdom. But if the Bank insists, and its
rich members concur, that a First World vision should be
imposed on a developing world and that the poor must be
elevated whether they live in countries that cannot afford
to pay or in countries that do not want to pay, it needs a
new financial structure to match modern realities.
Six World Bank Pretexts for Lending
The Bank wants to remain “relevant” to the middleincome countries but its defense of lending as a means
to “stayed involved” is rooted in the past and now refuted
by the facts.
I. The Bank lends to countries without ready
access to the capital markets
It is widely believed that the World Bank devotes the greater
part of its effort to countries denied market financing. In
truth, the Bank centers its portfolio on the most creditworthy candidates, a broad overlap with the private sector
that is specifically enjoined in its mandate.
A review of the Bank’s lending over the last five years
reveals that 99 percent went to countries with international
bond ratings from an investment-grade A down to a highyield/higher risk B. Approximately 25 percent of resources
flowed to nations with an investment grade rating and an
additional 74 percent to countries with high-yield ratings
at the time of the loan. More disquieting, the share of
IBRD loans to countries without international ratings has
fallen from 40 percent in 1993 to 1 percent in 2001–2005.
(See graph I.)
The Bank has contrasted the private sector’s
70 percent concentration of flows to 10 countries with
its own lending spectrum that channels resources to the
entire developing world. However, a review of the major
recipients of the IBRD’s resources over the last five years
reveals that 10 countries accounted for 76 percent of
flows, while the remaining 69 borrowing members were
left to divide only 24 percent.
These are the very countries that attract the bulk of
private sector resources: Turkey (14 percent), Brazil
(13 percent), India (10 percent), Mexico (9 percent), China
(8 percent), Argentina (8 percent), Colombia (5 percent),
Indonesia (3 percent), Romania (3 percent) and Russia
(3 percent). (See chart I.) There are another 17 emerging
nations with reliable access to the capital markets.1 Added
CGD0502 0527_Engl_6x9.indd 119
8/17/06 2:55:17 PM
120
Rescuing the World Bank
together, just 27 economies monopolize 90 percent of
Bank lending.2
In total, the 10 major borrowers received $2 billion in
net resources from the Bank over the past five years, or
an insignificant 0.4 percent of the $580 billion originating
in private sector medium- and long-term external debt,
portfolio equity and direct investment.3 (See chart I.)
II. The Bank lends where the developing world’s
poor live
As emerging nations have gained increasingly greater
access to financial markets, the Bank has conjured
up an alternate argument. It is the relative share in the
developing world, whether by population, by economic
size or by poverty that justifies the concentration of its
lending in so few major emerging countries.
The numbers deny this claim. Six of the Bank’s
10 leading clients annexed 52 percent of Bank loans over
the past five years, yet only accounted for 10 percent of
the total population and 24 percent of the GDP of IBRDeligible borrowers. Their average per capita income of
more than $8,000 on a purchasing power parity basis
placed them in the top quarter of emerging nations.
(See table I.)
III. The Bank lends for projects without interest
to the private sector
A host government guarantee is required on all Bank loans.
This displaces private sector lending dollar-for-dollar for
any country with capital market access and renders the
destination of proceeds irrelevant. When private lenders
can look to the host government for repayment, as the
Bank does, the capital markets are indifferent to end-uses.
Whether the goal is financing vaccinations of Indians in the
Amazon or nuclear weapons, prospectuses of sovereign
bond issues simply state “general government purposes”
as the use of proceeds.
IV. Bank lending is the sole source of funds for
long-term development
The benefits of many projects accrue over long-term
horizons and were once difficult to finance, even
for countries with ready access to the markets. Now,
the capital markets supply 20-year, 30-year and even
40-year financing, far beyond the Bank offer of amortizingloans with final maturities of 15–20 years. During the
CGD0502 0527_Engl_6x9.indd 120
8/17/06 2:55:17 PM
121
Selected Essays
past five years, 23 emerging World Bank borrowers have
issued bonds in the market with maturities stretching
into the future 25 years and more, well above the limits
of Bank terms.
V. Bank loans are a counter-cyclical balance to
volatile market flows
The specter of a sudden exodus by the private markets
in times of financial stress in contrast to the loyal and
steady flow of Bank funding is a timeworn argument.
But counter-cyclical stabilization requires more resources
than the Bank can muster. If private flows were to collapse
by 50 percent, Bank loans would still represent less
than 1/50th of the total capital moving into middleincome economies.
The global marketplace is remarkably resilient. Within
three months of the 1997 Asian crisis, Korea obtained
$4 billion in the capital markets with medium-term
maturities. When Brazil faltered in 1998, the next three
months counted 20 issues totaling $12 billion for
Latin American sovereign borrowers with maturities of
5–20 years. There was even a $2 billion issue for
Brazil itself.
A constant stream of lending in anticipation of the off
chance of a temporary fall in private sector flows cannot
be justified. When and if a crisis threatens, official funds
can be mobilized, but this is outside the mandate of a
development institution.
VI. Bank lending to emerging countries
generates profits to subsidize poor nations
Regardless of credit risk, the Bank charges the same
interest rate to all borrowers, equal to the Bank’s own
cost of funding in the capital markets, plus a spread of
0.50 percent per annum.4 This spread, when added to
commitment and up-front fees, is claimed to cover the
administrative costs of running the Bank and make a
profit that is passed on to the poorest nations. Far from
generating a surplus for the poor, lending is draining
resources. When all is accounted for, the Bank loses
money on its loans. For the past 12 years, annual deficits
of $100–500 million have resulted in a cumulative loss
from lending of $3.2 billion. (See graph II.)
In truth, the Bank earns its net income by using its
$40 billion of zero-cost capital.5 This pool of cash
generates between $1.5 billion and $2 billion per annum,
CGD0502 0527_Engl_6x9.indd 121
8/17/06 2:55:17 PM
122
Rescuing the World Bank
depending on the level of interest rates. And this net income
is the same whether it is lent to Bank borrowers or simply
placed in a portfolio of 10-year U.S. Treasury notes.
The record reveals that the Bank’s operating income
is a function of the level of interest rates. It closely
tracks the return on the Bank’s zero-cost capital invested
at the 10-year U.S. Treasury rate, adjusted for other
income and expense. (See graph II.) If the operating
income were derived from the spread on Bank loans, the
return would be constant, whatever the fluctuations in
U.S. Treasury rates.
Bundling Advice and Loans:
An Outdated Business Model
Doing business with the Bank is not just about
money. Lending has always been a two-part package.
There is a loan at highly subsidized rates, historically
7–10 percent per annum below the market. Clearly a gift.
And then there is the “technical assistance” which the
Bank insists is highly valued and the very reason clients
borrow from the Bank. In short, another gift. Yet the
Bank contends that borrowers will not follow the advice
unless it is partnered with subsidized loans.6 At first
hearing, all this defies logic and common sense. If the
Bank’s advice is truly “assistance,” why do borrowers
insist on being paid to comply?
Translated, this Bankspeak is really about imposing a
First World social vision upon an emerging world intent
on growth. If the environment must be safeguarded, if
workers must be protected, if women must play an equal
role, if indigenous peoples must be empowered and if
the overriding focus must be on the poor, the trade-off
has a cost.
Bundling really means that emerging nations are being
paid to execute projects low on national priorities, and
that they are being paid to implement projects in a manner
that imposes large costs, not just in money but in time
and effort, that arms-length market funds do not demand.
For decades, finance ministers of developing countries
have sat at the table and listened—after all, they were
being paid millions of dollars per hour in subsidies to
attend the lecture.
But in the past decade, three independent trends
have converged to destroy the attractiveness of Bank
loans: international capital markets expanded and
CGD0502 0527_Engl_6x9.indd 122
8/17/06 2:55:18 PM
123
Selected Essays
became willing to take on the risks inherent in developing
economies; emerging nations became stronger as sound
policies elevated their credit status; and nongovernmental
organizations (NGOs) were invited by the Bank to
step inside the development aid process as anointed
spokespersons for civil society. There are now some
20,000 highly vocal NGOs with a multiplicity of demands
that have led the Bank to slight the infrastructure needs
high on borrower plans for growth in favor of “social
programs” without clear economic yield, and to impose
elaborate standards that raise the cost of compliance
beyond practicality.
Bank “Advice” Has a Negative Value
Do borrowers come to the Bank for the advice or for the
subsidy? Now the facts are in and the debate is ended.
Since 1999, the subsidy in World Bank loans to major
emerging market governments has fallen from an average
12 percent per annum to less than 2 percent per annum,
as measured by the JP Morgan Emerging Market Bond
Index.7 At the same time, the net borrowing by these
nations from the Bank has collapsed from a positive
$14 billion in 1999–2002 to a negative $15 billion in
2003–2005. (See graph III.) First, borrowing slowed;
then countries moved on to repay loans. The interest
rate differential is no longer enough to persuade finance
ministers to realign their economic priorities with the
social agendas of the Bank’s rich members.
When it all adds up, the Bank’s “technical assistance”
has a negative value to its traditional client states. A
new generation of government officials, with PhDs from
MIT and Chicago, has done the arithmetic. Borrowing
patterns reveal that they rated the cost of Bank “advice” at
3–4 percent per annum. Over time, that amounts to
25–35 percent of loan expense. When the interest
subsidy fell below the cost of World Bank compliance,
the real subsidy vanished and so did the borrowers.
(See graph III.)
The conclusion: For years, World Bank loans have
been funding projects that countries didn’t think were
worth financing out of their own resources or worth the
cost of borrowing at a market interest rate. Borrowers
are willing to pay the markets 3–4 percent more as the
price of independent choice.
CGD0502 0527_Engl_6x9.indd 123
8/17/06 2:55:18 PM
124
Rescuing the World Bank
The Bank Is No Match for the Markets
The Bank was created to fill a void in the international
financial system. Resources and a willingness to assume
risk were needed to fuel growth in the developing world.
The private sector has now preempted the Bank’s role
of financial intermediary to emerging nations and is far
better at it than the Bank can ever hope to be. Yet the
Bank is clinging to its past.
Its traditional tools can no longer deliver the subsidy
that keeps it in the game. There is little wiggle room in the
0.50 percent annual charge, 0.25 percent commitment
fee and 1 percent up-front fee the Bank adds to its cost
of raising money to cover its own expenses. When all
costs are counted, the Bank is already losing money on its
lending. Cutting down on the burdens of the bureaucratic
“hassle factor” will have a minimal impact on the “price”
of its loans.
To counter the competitive threat of the private sector,
the Bank is ready to abandon the protections that have
served it well for decades and to search for innovative
financial instruments in the marketplace. But the effective
cost of Bank resources to its clients can never be lower
than its own cost of funds.
How to lower lending rates? How to assume more risk?
How to invent new instruments? These are all the wrong
questions. Abandoning the sovereign guarantee, lending
to sub-national entities, substituting guarantees for loans,
securitizing pools of loans and adding what Wall Street calls
the “nuclear waste” to the Bank’s portfolio. These are all the
wrong answers. Even the most convoluted mechanisms
to embed subsidies in new instruments will only lead to
hidden but ever-increasing costs for the Bank. And the
only outcome will be a growing exposure to risk without
compensation to cover losses.
The Bank is rational to consider risk lightly for it is well
placed to hide failures in ways that might put a private
sector institution out of business and its management
behind bars. But the Bank has no regulators. Its skills
of concealment were honed in the poorest economies
where, for two decades, a system of “defensive lending”
miraculously matched the dates and amounts of
repayment schedules to “new” loans, creating a perpetual
roll-over of defaulted debt. In the end, it was the Bank’s
rich members that assumed the losses of “debt relief”
and restored the Bank’s balance sheet.
CGD0502 0527_Engl_6x9.indd 124
8/17/06 2:55:18 PM
125
Selected Essays
When the Bank steps out of its protected bailiwick, it
is skilled private investors who will profit from the Bank’s
learning experience. Failures to make allowance for risk
in a futile effort to defend market share will be quietly
covered up. Bad loans will be hidden and rolled over.
Effective resources will be diminished. An invoice for the
losses will again be delivered to G7 taxpayers.
The Irrelevance of Lending
The Bank is no longer in a world short of capital. World
Bank lending is clouding the landscape and wasting
resources. All that the Bank really contributes in a world
of sophisticated financial markets is the subsidy that
fills the gap between the real cost of projects and what
recipients are willing to pay.
As the ratings of middle-income countries climb
and their cost of market funds falls, the Bank is being
forced to seek the help of other donors to recreate the
subsidy once provided by its loans. A pioneering project:
The Bank is building schools in China’s impoverished
Western provinces but the bill for interest charges is being
mailed to the United Kingdom, attention Chancellor of
the Exchequer Gordon Brown.
China is awash in money. There are $700 billion in
foreign reserves stored at its central bank and foreign
direct investment adds $60 billion each year to the
economy’s resources. Because the government can
borrow in the markets at a lower cost than from the Bank,
and because the Bank is more intent on aiding China’s
poor than China, the U.K. Treasury agreed to pick up
the interest tab on the China loans. In 20 years, when
China has paid back three loans totaling $300 million,
its cost will have been 55¢ on the dollar. All that China
really received and wanted was $12 million in annual
subsidies, not $300 million in loans.
If poor children are benefiting, where’s the harm? There’s
no harm if global aid resources are infinite. But the Bank’s
effort to retain influence with middle-income countries
siphons off scarce funds from the poorest. There is also
potential for harm if Bank loans free up prospering nations
to pursue other ambitions, perhaps nuclear weapons or
locking up access to natural resources abroad. (Iran has
been the Bank’s 10th largest borrower and China the
3rd largest over the last three years.)
CGD0502 0527_Engl_6x9.indd 125
8/17/06 2:55:18 PM
126
Rescuing the World Bank
If the Bank insists that the poor must be elevated
whether they live in countries that cannot afford to pay
or in countries that do not want to pay, if it wishes to
promulgate costly programs that are of marginal interest
to borrowers in the name of a freely interpreted version
of global public goods, more and more donor funds will
be required to restore the subsidy in Bank loans.
An unsustainable business model must be replaced with
a new financial structure that matches modern realities.
Lending is a blunt and inefficient instrument. The price of
persuasion should be at lowest cost. Subsidies can be
individually tailored according to the market borrowing
cost of governments and the priority the government
places on each project.
There is already $40 billion of zero-cost capital on
the Bank’s balance sheet as a starting point to endow
a permanent foundation that would be invested in the
capital markets to generate a stream of subsidies. These
would underwrite interest payments on country borrowing
in the markets. Over time, rich countries may be asked
to contribute more funds.
Otherwise, as more emerging nations move up the
credit ladder, donors will be compelled to divert an everincreasing share of their own aid funds to enhance the
appeal of Bank loans. The experiment begun in China
will be the prototype as Mexico, Brazil, Chile and others
line up for the same deal.
Mechanics should not be confused with the mission. The
Bank must accept that it is in the development business,
not the banking business. Long ago, they may have been
one and the same, but now there are better ways to deliver
resources to what the Bank perceives as its real clients,
the global poor, and to foster global public goods. If the
Bank continues to fight the tape, it will become irrelevant
to the purpose for which it was designed.
CGD0502 0527_Engl_6x9.indd 126
8/17/06 2:55:18 PM
127
Selected Essays
CGD0502 0527_Engl_6x9.indd 127
8/17/06 2:55:19 PM
128
Rescuing the World Bank
Chart I
Insignificant to its Major Clients:
World Bank versus Private Sector: Net Flows 2001-2005
World Bank 0.4%
Ten Leading World Bank Borrowers
Turkey
Brazil
India
Mexico
China
Argentina
Colombia
Indonesia
Romania
Russia
14.3%
13.0%
10.2%
8.8%
8.2%
7.5%
5.4%
3.3%
2.9%
2.8%
Total
76.4%
Private Sector 99.6%
Sources: World Bank, Global Development Finance 2005
3
CGD0502 0527_Engl_6x9.indd 128
8/17/06 2:55:20 PM
129
Selected Essays
Table I
Six Leading World Bank Borrowers:
Bank Loans versus Population, GDP, and Poverty
% of IBRD Loans
2001-2005
% of IBRD-Eligible % of IBRD-Eligible
2003 Population
2003 GDP
2003 Per Capita
Income (PPP)
Turkey
Brazil
Mexico
Argentina
Colombia
Romania
14.3%
13.0
8.8
7.5
5.4
2.9
1.6%
4.1
2.3
0.9
1.0
0.5
3.0%
7.2
9.6
2.1
1.2
0.7
$6,710
7,510
8,980
11,410
6,410
7,140
Total/Average
51.9%
10.4%
23.8%
$8,027
Sources: World Bank, World Development Indicators 2005
4
CGD0502 0527_Engl_6x9.indd 129
8/17/06 2:55:21 PM
130
Rescuing the World Bank
Graph II
Source of World Bank Income: Zero-Cost Capital, Not Loans
3,500
3,000
Reported Operating Income
Millions of U.S. Dollars
2,500
2,000
1,500
1,000
Net Income on Zero-Cost Capital*
500
00
-500
Net Income on Loans**
-1,000
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
*Estimated based upon the 10-year US Treasury interest rate and adjusted for changes in provisions for loan losses, contributions
to special programs, net loan income, arbitrage income and other income to ensure comparability to reported operating income.
**Interest spread income plus commitment fees plus up-front fees on IBRD loans minus IBRD administrative expense.
Sources: World Bank , Federal Reserve Historical Data
5
CGD0502 0527_Engl_6x9.indd 130
8/17/06 2:55:22 PM
131
Selected Essays
Graph III
20
12.00
16
Argentina Crisis
10.00
12
Brazil Crisis
8
8.00
Interest Subsidy
6.00
4
4.00
0
Net Borrowing
Ju
l-0
5
05
Ja
n-
Ju
l-0
4
4
Ja
n0
Ju
l-0
3
3
Ja
n0
Ju
l-0
2
2
Ja
n0
Ju
l-0
1
Ja
n-
Ja
n0
Ja
n9
01
-8
Ju
l-0
0
0.00
0
-4
Ju
l-9
9
2.00
Net IBRD Borrowing by Major Emerging Countries (FY; $billion)
14.00
9
Interest Subsidy in IBRD Loans (% per annum)
World Bank and Major Emerging Countries:
Falling Subsidies; Falling Loans
Sources: World Bank, JPMorgan
6
CGD0502 0527_Engl_6x9.indd 131
8/17/06 2:55:23 PM
132
Rescuing the World Bank
Notes
1. Bulgaria, Chile, Costa Rica, Ecuador, Egypt,
El Salvador, Korea, Malaysia, Morocco, Peru, Philippines,
Poland, Slovak Republic, Thailand, Tunisia, Uruguay and
Venezuela.
2. Within the middle-income group, there are 52 mostly
small economies that may rely on official financing in times
of stress. Though significant in number, these nations
received only 9 percent of Bank loans over the past five
years and account for only 8 percent of developing world
population.
3. These figures underestimate the quantity of private
sector inflows because the substantial foreign investment
in domestic bonds is not included in the data.
4. The interest spread charged on Bank loans was
0.25 percent in 1994–1998, 0.45–0.50 percent in 1999–
2000 and 2002–2005, and 0.35–0.50 percent in 2001.
In addition, the Bank charges a commitment fee of 0.25
percent and, from 1999, instituted an up-front fee of
1 percent.
5. As of June 30, 2005, the Bank’s $39 billion zero-cost
capital was comprised of $11.5 billion in paid-in capital
and $27.2 billion of retained earnings.
6. Another justification for the bundling is that loans
are concrete proof of the Bank’s confidence in its own
counsel. But the sovereign guarantee on Bank loans
divorces project results from the risk of loss. Whether the
advice is good or defective, whether the project succeeds
or fails, the borrower must repay the Bank.
7. A composite of 19 leading emerging market
sovereign borrowers.
CGD0502 0527_Engl_6x9.indd 132
8/17/06 2:55:23 PM
The World Bank and
the Middle-Income Countries
by David de Ferranti1
T
he World Bank’s role in middle income developing
countries needs to change. Not to end lending to
them, or adopt the other proposals from extremists
on the right or left. But rather to modernize both what
the Bank does and how it does it, so as to respond more
effectively to the changed circumstances, needs, and
preferences of this group of countries.2
Recommendations on how the Bank should modernize
are set out below. First, though, the case for it to stay
engaged is discussed, since a handful of voices are still
trying to argue otherwise.
The World Bank should remain engaged in the
middle-income countries
Arguments for axing World Bank lending to middle-ranking
developing countries enjoyed short-lived notoriety a few
years ago, with the publication of a report by Prof. Alan
Meltzer.3 Since then, however, that fringe view has been
endorsed only by a handful of American conservative
academics (primarily those who worked on the report
in the first place).
Few know this better than Paul Wolfowitz. Nominated
in 2005 as the new President of the Bank by a strongly
conservative US administration, of which he had been
a key member, Wolfowitz’s appointment was initially
acclaimed by the critics on the right. (“An inspired choice,”
wrote Alan Meltzer in The Wall Street Journal on March
18, 2005.) But Wolfowitz didn’t fall for their odd theories.
In his first Annual Meetings speech in September 2005
he stated unequivocally that “To help the middle income
countries grow and prosper, we need to continue to tailor
David de Ferranti, Brookings Institution
David de Ferranti was Regional Vice President for Latin America and the
Caribbean at the World Bank until his retirement in 2005. Currently he has
positions at the Brookings Institution and the United Nations Foundation,
and is the Chair of the Board of the Center on Budget and Policy Priorities.
Earlier, he held senior positions in the US government and directed research
at Rand. His research and writing have examined both developing country
and US domestic issues, and he is an adjunct professor at Georgetown
University.
CGD0502 0527_Engl_6x9.indd 133
133
8/17/06 2:55:23 PM
134
Rescuing the World Bank
our knowledge and financing to their specific needs.”4
Subsequently, on the eve of a visit to Brazil, he was quoted
as saying, “I really want to underscore the World Bank’s
commitment to Brazil and all the other middle income
countries in Latin America...”5
Nor is Wolfowitz alone. The Bank’s 184 shareholder
governments—liberal, conservative, and everything inbetween—have had numerous opportunities to review and
re-decide the Bank’s engagement in the middle income
countries. Instead of embracing the terminate-lending
schemes, they have repeatedly come down firmly, and as
a rule unanimously, on the side of continuing the Bank’s
important development work—analytical, operational and
financial—in this critical group of countries.6
Watch out for the spin....
The tiny band of diehards have not helped their case by
“spinning” the facts through the use of carefully selected
statistics. Here are a few examples.
They claim that IBRD loan demand has collapsed. The
truth is different. Figure I below gives the facts: IBRD
lending commitments each year over the past 15 years.
Lending shows significant fluctuation. It shot up during
1998 and 1999, when the Bank participated in several
crisis assistance packages. Levels then fell back, and for
a while were appreciably below those of the early to mid
Figure I
CGD0502 0527_Engl_6x9.indd 134
8/17/06 2:55:24 PM
135
Selected Essays
1990s, in considerable part due to a premature cut-back in
Bank lending for infrastructure, based on overly optimistic
assumptions about the private sector’s readiness to
pick up this financing responsibility (the infrastructure
retrenchment is one the Bank has just recently begun to
reverse). The level in the latest completed year (fiscal year
2005) was some 6–7 percent down on that immediately
before the 1998–99 crisis. Looking forward, lending in
the first half of fiscal 2006 significantly surpassed lending
in the same period of fiscal 2005.7
How do the spinners transform this rather mundane
picture into an Emergency Room? Step one: start the
comparison from an atypical base—in this case, kicking
off from 1999’s record lending. Step two (and more
importantly): compare apples and oranges, by mixing
up lending with the pre-payment of older IBRD debt.
Like many US homeowners, some IBRD borrowers took
advantage of recent record low interest rates to refinance
their older, higher-interest debt, assuming the opportunity
would not last for ever. This is no more an indicator
of demand for future IBRD lending than homeowners
refinancing their mortgages signals the collapse of the
home loan market. In short, if Mark Twain had seen
this claim of “collapse”, might have been reminded of
his remark on hearing that the New York Journal had
published his obituary, “An exaggeration.”
Another example is the assertion that it is “disquieting”
that IBRD lending to countries without international
ratings has fallen from 40 percent in 1993 to 1 percent
in 2001–05. What this misleading statement obscures is
that the number of countries without a credit rating has
itself shrunk enormously over the period in question, as
more and more countries have sought out ratings. So,
a country without a rating is today almost an oddball.
Among borrowers from IBRD during the past five years,
only 7 unrated countries remained (Algeria, Belarus,
Uzbekistan, and four small Caribbean island nations)—
they incidentally accounted for less than 1 percent of the
IBRD poor. A further dozen non-rated countries were
non-borrowers from IBRD, due either to the absence of
a supportable program or to having been in “non-accrual
status”—i.e., not up-to-date in servicing their debts to
the Bank—such as Zimbabwe, for example. The critics’
disquiet thus looks more than a trifle overdone.
CGD0502 0527_Engl_6x9.indd 135
8/17/06 2:55:24 PM
136
Rescuing the World Bank
Still another example is the claim that IBRD lending
largely by-passes the countries where the poor live. In
fact, the top 10 borrowers from IBRD over the past five
years, mostly among the largest countries, together
accounted for about 84 percent of all the poor people
(under $2 a day) living in the MICs as a whole (a further 5
percent of the MIC poverty was accounted for by Pakistan,
an important World Bank Group “blend” borrower, but
one that largely borrows from IDA). (See Table I.) Even if
one cherry-picks the list, as the critics sometimes do, to
remove the four big borrowers with the largest numbers of
poor (China, India, Russia, and Indonesia), the remaining
six countries still accounted for about 22 percent of the
IBRD poor living outside the four giants and got just over
50 percent of the lending. Beyond this, there can be good
reasons for IBRD support even in countries that are not
among those with the most poor people. A country in the
midst of a crucial reform program—such as some of the
former Soviet bloc countries—might want and need help,
and the world (and their poorer country neighbors) might
be better off if they got it. Overall, though, IBRD lending
comes much closer than the spinners acknowledge to
matching concentrations of dire poverty.
Juggling the data also hides something much more
important. When the options and their pros and cons
are even-handedly examined in balanced, reasoned
debate, there are compelling, broad-based reasons
why it makes sense for the Bank to stay engaged in the
middle income countries. Extensive work has been done
examining the reasons.8 A later section here outlines that
terrain, reinforcing the conclusion that the Bank should
stay engaged. Prior to that, the real aim of this chapter
takes center stage: how should the Bank improve?
How the Bank should modernize its work in
the middle-income countries
Modernize Financial Products
Borrowers report that, while the Bank’s traditional loans
may have once been appropriate, the institution now
needs to realize that new and different instruments may
be more responsive to their needs. These arguments
need to be listened to.
The Bank has in fact significantly modernized its product
offerings. However, many of these new products do not
appear, at least until very recently, to have been promoted
CGD0502 0527_Engl_6x9.indd 136
8/17/06 2:55:24 PM
137
Selected Essays
very actively. Whether one is talking about guarantees,
lending in local currency, or insurance products—or the
possibility of lending to sub-national levels of government
without necessarily requiring sovereign guarantees—
more could be done.
The Bank should also look seriously at recent advances
in financial markets. This includes “structured finance”
approaches where its participation could leverage in far
more capital—tapping much more of the huge private
sector potential to help development—than is possible
through old-style, go-it-alone projects. Proven products
take a diversified pool of investments, unpack the risks,
and repack them into different tranches matching the
risk/reward appetites and capabilities of different classes
of investors. The Bank should review whether it should
take positions in these areas. In addition, there may
be other, perhaps better options out there for vehicles
whereby the Bank could leverage greater flows from the
private sector.
Cutting Down the Hassle
Many observers—and especially borrowers—feel that
the steps and requirements that must be complied with
to obtain a Bank loan are still crushingly burdensome,
despite recent efforts to lighten the load. The Bank Group
needs to take a new look at this “hassle factor.”
To some degree, these demands represent a prudent
concern to ensure due diligence. “Safeguard” policies,
in particular—in areas like a project’s environmental
impact or effects on local residents such as indigenous
people—largely reflect the lessons of experience, and
the need to take reasonable precautions. Yet there is
also the danger that, under the influence of single-issue
pressure groups, agencies like the Bank take refuge in
demanding ever-more studies.
The key point here is to make sure that the substance
of key risks is addressed—and suitable risk mitigation
strategies adopted. But, especially when dealing with
more sophisticated borrowers, the Bank should be more
willing to work with countries’ own national systems
of safeguards, where these achieve substantively
comparable protection to the Bank’s own procedures,
and should focus on “upstream” remedies of root causes
rather than downstream fixes to projects that are already
well advanced. Resistance to this approach by some
CGD0502 0527_Engl_6x9.indd 137
8/17/06 2:55:24 PM
138
Rescuing the World Bank
shareholder representatives suggests a failure to think
the issue through properly.
Learning from Differences across Countries
The Bank should review with some care—and aim to
learn from—the variations in its client relationships as
between one middle income country and another (and
one region and another). Some countries and regions
have shown continued strong demand for World Bank
products, in others interest appears to have weakened.
Are the differences inherent to the countries themselves?
May some of the differences reflect alternative strategies
the Bank has adopted across different regions
and countries?
Experience of working in Latin America, for example,
prompts the question of how far the Bank’s successful
efforts to appoint a substantial number of managers and
senior staff from within that region may have helped keep
the Bank relevant to borrowers’ needs. The ability to
identify with borrowers and their culture—and speak their
language, literally and figuratively—may be one key to
staying relevant.
Performance-Based Lending
The argument for providing more support on a
“performance-based” basis is compelling. The basic
concept is simple. Rather than financial flows being
triggered by a country’s “inputs” (such as its own
spending on health), the performance model ties funding
to “outputs” or performance indicators, such as the
number of children immunized.
The main issues are practical, not “ideological”: how to
set meaningful performance indicators, establish reliable
systems for monitoring them, make sure no essential
components get missed out (such as focusing so heavily
on “new” coverage that one neglects to measure upkeep
of existing systems). They are not easy challenges, but
they should be tackled.
Loan Terms
Some commentators have proposed considering further
differentiation of loan terms for different countries. One
line of argument calls for stronger, richer countries to pay
more, since they are better able to pay. Another makes
the converse argument—that the less creditworthy should
pay more because they are a worse risk. Elements of
CGD0502 0527_Engl_6x9.indd 138
8/17/06 2:55:24 PM
139
Selected Essays
both arguments are in fact embedded in current pricing
policies. The difference between IDA terms and IBRD
terms applies the first argument—the poorer pay less.
The harder-than-normal terms adopted for “special”
lending—under emergency conditions—requires riskier
lending to carry a higher price.
Both Bank officials and the critics agree that current
IBRD lending terms are hardly softer (if at all) than those
the best-rated borrowers can obtain from the markets.
In addition, IBRD loan spreads over the Bank’s cost of
borrowing are already reckoned to more-than-cover the
direct costs of the Bank’s “banking” business and to make
a hefty contribution to the cost of such “public goods”
functions as research and analysis. One might ask how
much more of these overhead costs should reasonably
be included directly in loan charges.
The trump card in this debate is that the Bank generally
revises its basic policies only on the basis of a broad
consensus among the shareholders. And consensus on
further change in this area will prove hard to come by.
Nevertheless, the shareholders have a responsibility to
try their best to overcome differences between them, and
thus should ask for a systematic look at the issue.
Expanding Intellectual Partnerships
Finally, while the Bank has definitely come some way in
combating the “not invented here” syndrome, there is
still a way to go. Experience suggests that the Bank still
under-uses intellectual capacities outside the institution.
There has been an explosion in the numbers of highlytrained professionals in many borrowing countries, and
in the capacity of domestic think-tanks, consulting
firms, research institutions, and university departments.
There is still room for more analytical work to be done
in partnership with local organizations. This can benefit
both sides—building local capacity further, and improving
the quality of the analysis by incorporating different
perspectives. A Bank that partners more with others—in
earnest and not just in rhetoric, and draws on (and scales
up) ideas developed by others—might also be a Bank
that does not need as many staff and as big a budget for
them as it would otherwise. Certainly, the composition
of the staff would need to change, all the more so if the
other recommendations here were adopted, especially
the one on financial products.
CGD0502 0527_Engl_6x9.indd 139
8/17/06 2:55:24 PM
140
Rescuing the World Bank
Other actions too have been widely proposed that would
help, including some relating to the composition, role,
and budget of the Bank’s Board of Executive Directors,
and others on improving evaluation of Bank operations.
There is not space here to go into all of them,9 but one
overarching point is fundamental.
Modernizing the Bank thoroughly will require
contributions by everyone—its President, managers,
staff, and external groups, but especially by its member
country governments themselves, both through their
positions on the Board and at the higher levels where
major global policy choices are decided. For too long,
the vital role of the member countries’ leading officials
and representatives in determining what the Bank can
be and do—and the impossibility of bringing about major
change in the Bank without their active leadership—have
been greatly under-recognized, especially by those not
extensively familiar with the inner workings of the Bank.
And for too long too, member countries’ leaders have
failed to find ways to grapple effectively with some of
the biggest and toughest questions about the Bank
and its future, including the question of its role in the
middle income countries. Piecemeal efforts on selected
issues—for example, on the low-income countries and
on debt reduction—and through periodic discussions
in the G8 and other fora, have achieved notable gains,
but also created troublesome inconsistencies. A more
thorough grappling with core issues, however hard
politically, and however long it may take to be fruitful, is
of urgent priority.
More on why the Bank should stay engaged
Returning now, as promised, to the case for a continuing
Bank role in the middle income countries, there are
several parts to the story, including the answers to two
basic questions:
• Should the larger world community—the Bank’s
shareholders—care about developments in the middle
income countries and try to influence them?
• Assuming they do, should they work through official
development agencies like the World Bank, rather
than leaving the job to market forces and/or making
ad hoc institutional arrangements?
CGD0502 0527_Engl_6x9.indd 140
8/17/06 2:55:25 PM
141
Selected Essays
To answer the first question positively—as governments
around the world have in fact done resoundingly—involves
recognizing that we live in an increasingly interconnected
world, where developments on the other side of the
globe can affect our economic well-being, our health,
our security and the global environment our grandchildren
will inherit. Old dreams of isolationism look threadbare
in a world of globalized production, finance and trade,
international terror threats, pandemics like HIV/AIDS and
bird flu, and global environmental challenges like loss of
biodiversity and climate change.
Indeed, what happens in the middle income countries
matters a lot in the global picture:
• The MICs account for around two thirds of the world’s
total population. Their economies, meanwhile, provide
important and growing sources of export demand for
the wider world’s producers and of potential investment
opportunities for other countries’ investors.
• The MICs include roughly three quarters of all the
people living in poverty (under $2 a day) around
the world.
• The MICs are now big enough to create systemic
risk in global financial markets. A high proportion of
recent global financial crises have originated in MICs
like Mexico, Russia, East Asia, Turkey and Brazil.
• On strategic issues, MICs repeatedly emerge as key
players (the aftermath of the break-up of the Soviet
empire, the turmoil in the former Yugoslavia, tensions
in the Middle East and South Asia, etc., etc.).
• MICs account for an estimated 47 percent of global
CO2 emissions.
• MICs account for over half the world’s areas protected
for their environmental significance.
So, why then work through the Bank? A modernized,
well-functioning Bank, as imperfect as it will always be,
can be shaped into the best instrument that the world’s
countries are likely to have in the foreseeable future for
helping achieve at least some of their global objectives.
Among its relatively unique combination of attributes for
this role are:
• broad-based analytical expertise on development policy
issues at the global, regional and national levels;
CGD0502 0527_Engl_6x9.indd 141
8/17/06 2:55:25 PM
142
Rescuing the World Bank
• the ability to combine an appreciation of the broad
macro perspective with detailed examination of
policy issues at the sectoral and micro levels, and
a proven capacity to take on new challenges;
• extensive operational experience in implementing
reform and investment programs in different
geographical and sectoral contexts; and
• sufficient financial capacity to be able to match its
intellectual contribution with resource commitments
that reinforce its partnership with members
throughout the implementation phase.
At the heart of the critics’ case, though, is the relationship
between the World Bank and private capital markets.
Repeatedly, they come back to this comparison: lending
by the Bank, they say, necessarily crowds out lending by
the markets, lending by the Bank is pitifully tiny compared
to the scale of the markets, the Bank cannot compare
with the efficiency of the markets, the Bank should not
lend to countries with access to the markets....
None of this is new. Those who know the Bank expect
criticism from both ends of the political spectrum. Critics
on the far left accuse the Bank of being a tool for the spread
of international capitalism. Those on the right complain
that it is not. Of the two, the leftists seem to have the
better factual grasp of what the Bank actually does.
Missing the point on public-private
complementarity....
Missing from the conservative critiques is any sense
of the importance of complementarity between public
agencies and private markets. To the critics, any public
lending to a country with market access must of necessity
supplant private lending dollar for dollar—they see a
“zero-sum game”. Yet most economists today recognize
that efficient private markets do not appear magically,
but require supporting public infrastructure, institutional
as well as physical. And much of what the World Bank
actually does directly helps to improve the climate for
private investment:
• The Bank has encouraged and supported countries
in implementing trade reforms to open up to greater
international competition, and in removing restrictive
regulations on inward foreign direct investment.
CGD0502 0527_Engl_6x9.indd 142
8/17/06 2:55:25 PM
143
Selected Essays
• In utilities and infrastructure, the Bank has very
actively promoted expanding private provision.
• The Bank helps clients strengthen the essential
legal and judicial infrastructure for private markets,
including the regulatory frameworks that underpin
competitive private financial markets It also helps
countries confront corruption, which—among its
other evils—distorts the “level playing field” needed
by efficient markets.
• The Bank’s work on national regulatory frameworks—
including its annual published comparisons of “Doing
Business” in some 155 countries—provide powerful
advocacy tools in favor of freeing business from
harmful and superfluous regulations.
• The Bank works alongside other agencies, like the
IMF, to help countries emerge rapidly from macrofinancial crises when they have temporarily lost the
confidence of the private markets. Complementing
the IMF’s focus on rectifying macroeconomic
imbalances, the Bank’s emphasis is on promoting
crucial structural reforms and protecting vulnerable
social groups.
• The Bank’s work in helping countries improve the
education and health of their populations, and
upgrade basic infrastructure, provides crucial
support for future market-driven development.
Even the most committed advocates of market-driven
development may find it hard to object to most of these
efforts—which may incidentally explain why the critics
devote so little of their prolific output to discussing what
the Bank actually does.
Deconstructing the Bank?
A fall-back for the critics is to argue that, even if what
the Bank does might not be 100 percent objectionable,
the institution itself is superfluous. Everything the Bank
does, they say, could be picked up by the private sector.
Private markets could lend where the Bank lends (or at
least in the more creditworthy countries), and consulting
firms could provide any technical advice needed.
At the theoretical level, one can argue for breaking
up any complex organization. Why not replace our
cumbersome universities by independent tutors, as in the
middle ages? Private certification bodies could compete
CGD0502 0527_Engl_6x9.indd 143
8/17/06 2:55:25 PM
144
Rescuing the World Bank
to provide qualifications. College football teams could
be sold to the NFL....
As with universities, the case against breaking up the
World Bank involves recognizing that “the whole is greater
than the sum of the parts”. The Bank’s global reach,
operational involvement and financial strength enable it
to serve an unparalleled “global public goods” function
as a respected world center of practical development
experience, data and information.
Still, why “bundle” technical inputs with finance?
Why not just provide technical advice and let countries
go to the markets for resources? Experience points to
three factors.
First, for many countries, access to the markets is
more problematic and variable than the critics admit.
They paint market flows as dwarfing official lending,
but most private flows go to private investments—car
factories, hotels, Cola bottling plants, etc. In aggregate,
average private lending for public (or publicly guaranteed)
purposes is roughly comparable in scale to the lending of
official agencies, including the Bank. But private lending
is far more subject to “sudden stops” in crisis times. And
for many borrowers, especially those without investment
grade ratings, the effective costs of private borrowing
can be steep.
Secondly, even if, in a perfect world, sound advice
would sell itself based on quality alone, in the real world,
the willingness to back substance with hard resources
can often be the price of getting through the door to
present one’s ideas in the first place.
Thirdly, the knowledge that the Bank is willing to
commit its resources to a program offers re-assurance
that it will not walk away from the borrower. We all know
jokes about consultants who turn in their report and then
respond “I don’t do implementation.” The Bank cannot
offer that excuse.
This does not imply that the Bank should never offer
advice without funding. Indeed it now provides feebased advisory services to a number of its clients. But
a distinction should be made. Analytical work that is
essential for maintaining the Bank’s “public good” role of
reporting on key development issues should continue as
part of the essential package of client services. Advice in
areas of very specific country interest, by contrast, lends
itself to being placed on an optional, fee-based basis.
CGD0502 0527_Engl_6x9.indd 144
8/17/06 2:55:25 PM
145
Selected Essays
Who should be able to borrow?
A key element in the public debate is very different views
on who should be eligible to borrow from IBRD. The
approach taken by the shareholders is summarized in
the Bank’s 2005 Annual Report:
“In fiscal 2005 countries with a per capita income of less
than $5,295 that were not IDA-only borrowers were eligible
to borrow from IBRD. Countries with higher per capita
incomes were able to borrow from IBRD under special
circumstances, or as part of a graduation strategy.”10
The Bank’s shareholders thus base eligibility primarily
on a country’s overall state of development (as proxied
by per capita income). They apply the approach with
some flexibility, allowing for a transition process and
for special circumstances, as when Korea temporarily
returned to IBRD borrowing status in 1997 (three years
after “graduating”), when it lost the confidence of the
markets during the wider East Asian crisis.
The critics proposed a very different approach
in the report of the majority group within the Meltzer
commission:
“All resource transfers to countries that enjoy capital
market access (as denoted by an investment grade
international bond rating) or with a per capita income
in excess of $4000, would be phased out over the next
5 years. Starting at $2500 (per capita income), official
assistance would be limited. (Dollar values should be
indexed). [For the record, indexation since 2000 would
raise the above dollar figures to roughly $4500 and $2800,
respectively, in late 2005 terms].”11
Meltzer’s proposal to arbitrarily limit lending to countries
with per capita incomes above $2800, and to apply a rigid
phase out of all lending to countries with income per head
of over $4500, would knock out or limit development
support to most developing and transition countries in
Latin America and Eastern Europe. It would convert the
Bank from a strong development agency with a global
reach into a much-shrunk body dealing primarily with
Africa and a few low-income Asian countries.
Meltzer’s addition of “market access” as a further
reason for withdrawing eligibility to borrow would be an
even more radical departure.12 A borrower’s access to
private lending, as measured by agencies’ credit ratings,
does not reflect its level of development, so much as its
prudence in borrowing and servicing its debts. India
CGD0502 0527_Engl_6x9.indd 145
8/17/06 2:55:26 PM
146
Rescuing the World Bank
undoubtedly deserves credit for the policy reforms that
recently lifted it to an “investment” rating. But with 850
million Indians (four in five of the population) surviving
on less than $2 a day, one may question whether the
international community truly wants its congratulatory
card to India to read, as the critics would draft it, “You’re
on your own now!”
The heart of the matter?
The critics have concentrated their fire on the World Bank.
But their central objections to IBRD lending to MICs apply
with comparable logic to any official development lending
to these countries—whether from regional banks, bilateral
development agencies or wherever. Their real objection
is evidently not to the specifics of the Bank’s lending
programs or its policy advice—subjects they barely begin
to discuss. Nor have they seriously tried to prove the
Bank less competent than its peers. Rather, the core of
their case—even if generally camouflaged beneath the
quibbling over this or that detail about the Bank—implies
hostility to public development work in and of itself. Like
left-wing activists who mobilize against McDonalds rather
than its less-conspicuous competitors, the critics have
identified the World Bank as the most visible symbol
of public development assistance—and opposition to
IBRD’s work in the middle income countries as the thin
edge of a larger ideological wedge.
Table I
The Middle Income Countries
C02
Protected
Emissions
Areas (Thousands
IBRD Eligible
CGD0502 0527_Engl_6x9.indd 146
(Thousands
of Metric
Countries1 of Hectares)2
Tons)2
Estimated
% under
Population
Population3
$2/day4
under $2/day
Algeria
11,864
74,176
32,531,853
15.10
4,912,310
Antigua and
Barbuda
0
359
68,722
NA
NA
Argentina
5,911
138,983
39,537,943
14.31
5,657,880
Azerbaijan
394
29,490
7,911,974
9.10
719,990
Barbados
0
1,334
279,254
NA
NA
Belarus
1,304
59,561
10,300,483
0.68
70,043
Belize
633
827
279,457
NA
NA
8/17/06 2:55:26 PM
147
Selected Essays
Table I (continued)
The Middle Income Countries
C02
Protected
Emissions
Areas (Thousands
IBRD Eligible
(Thousands
of Metric
Countries1 of Hectares)2
Tons)2
Estimated
% under
Population
Population3
$2/day4
under $2/day
Bolivia
12,082
11,714
8,857,870
34.30
3,038,249
Bosnia and
Herzegovina
27
14,269
4,025,476
NA
NA
Botswana
10,499
4,033
1,640,115
50.10
821,698
Brazil
32,866
327,858
186,112,794
22.43
41,745,100
Bulgaria
594
44,731
7,450,349
16.20
1,206,957
Chile
2,650
54,790
15,980,912
9.58
1,530,971
China
105,527
3,473,597
1,306,313,812
46.70
610,048,550
Colombia
9,786
63,998
42,954,279
22.56
9,690,485
Costa Rica
477
5,223
4,016,173
9.45
379,528
Croatia
339
19,191
4,495,904
0.53
23,828
Czech
Republic
196
124,069
10,241,138
0.23
23,555
Dominica
10
76
69,029
NA
NA
Dominican
Republic
1,113
19,887
8,950,034
0.76
68,020
Ecuador
2,308
20,705
13,363,593
36.09
4,822,921
Egypt,
Arab Rep.
4,536
127,131
77,505,756
43.90
34,025,027
El Salvador
NA
6,598
6,704,932
58.02
3,890,202
Equatorial
Guinea
455
716
535,881
NA
NA
Estonia
350
14,884
1,332,893
4.69
62,513
Fiji
16
701
893,354
NA
NA
Gabon
80
1,455
1,389,201
NA
NA
Grenada
NA
79
89,502
NA
NA
Guatemala
594
10,097
14,655,189
37.36
5,475,179
Hungary
821
56,850
10,006,835
1.52
152,104
India
15,291
1,007,979
1,080,264,388
79.90
863,131,246
Indonesia
8,607
286,027
241,973,879
52.42
126,842,707
Iran,
Islamic Rep.
10,376
297,930
68,017,860
7.30
4,965,304
Iraq
1
78,507
26,074,906
NA
NA
Jamaica
0
10,320
2,731,832
13.30
363,334
Jordan
913
15,535
5,759,732
7.40
426,220
Kazakhstan
7,742
123,686
15,185,844
8.45
1,283,204
CGD0502 0527_Engl_6x9.indd 147
8/17/06 2:55:26 PM
148
Rescuing the World Bank
Table I (continued)
The Middle Income Countries
C02
Protected
Emissions
Areas (Thousands
IBRD Eligible
CGD0502 0527_Engl_6x9.indd 148
(Thousands
of Metric
Countries1 of Hectares)2
Tons)2
Korea,
Republic of
350
Estimated
% under
Population
Population3
$2/day4
under $2/day
470,020
48,422,644
1.00
484,226
Latvia
818
6,490
2,290,237
8.30
190,090
Lebanon
4
15,569
3,826,018
NA
NA
Libya
122
42,275
7,765,563
NA
NA
Lithuania
592
11,574
3,596,617
6.90
248,167
Macedonia,
FYR
180
8,862
2,071,210
4.00
82,848
Malaysia
1,366
123,603
23,953,136
9.30
2,227,642
Marshall
Islands
NA
NA
59,071
NA
NA
Mauritius
7
2,796
1,230,602
NA
NA
25,807,305
Mexico
1,205
385,075
106,202,903
24.30
Micronesia
5
NA
108,105
NA
NA
Morocco
326
33,236
32,725,847
14.30
4,679,796
Namibia
3,214
1,945
2,030,692
55.80
1,133,126
Pakistan
3,509
105,983
162,419,946
65.60
106,547,485
Palau
0
242
20,303
NA
NA
Panama
483
5,709
3,039,150
17.90
544,008
Papua
New Guinea
7
2,445
5,545,268
NA
NA
Paraguay
1,391
3,659
6,347,884
30.29
1,922,774
Peru
4,010
28,194
27,925,628
37.71
10,530,754
Philippines
1,513
75,299
87,857,973
47.48
41,714,966
Poland
3,417
303,777
38,635,144
1.18
455,895
Romania
476
90,729
22,329,977
20.50
4,577,645
Russian
Federation
90,223
1,540,365
143,420,309
23.80
34,134,034
Serbia and
Montenegro
327
44,355
10,829,175
NA
NA
Seychelles
4
224
81,188
NA
NA
Slovak
Republic
357
36,927
5,431,363
2.40
130,353
South Africa
6,461
344,590
44,344,136
34.07
15,108,047
St. Lucia
2
446
166,312
NA
NA
8/17/06 2:55:27 PM
149
Selected Essays
Table I (continued)
The Middle Income Countries
C02
Protected
Emissions
Areas (Thousands
IBRD Eligible
(Thousands
of Metric
Countries1 of Hectares)2
Tons)2
Estimated
Population3
% under
Population
$2/day4
under $2/day
St. Vincent
and the
Grenadines
4
165
117,534
NA
NA
Suriname
1,846
2,244
438,144
NA
NA
Swaziland
35
388
1,173,900
22.55
264,714
Syrian Arab
Republic
NA
51,347
18,448,752
NA
NA
Thailand
6,516
171,697
65,444,371
32.50
21,269,421
Trinidad
and Tobago
24
18,090
1,088,644
39.00
424,571
Tunisia
28
20,179
10,074,951
10.00
1,007,495
Turkey
571
223,862
69,660,559
10.30
7,175,038
Turkmenistan
1,883
34,584
4,952,081
44.00
2,178,916
Ukraine
1,937
348,357
47,425,336
45.70
21,673,379
Uruguay
30
6,409
3,415,920
1.00
34,159
Uzbekistan
2,050
121,045
26,851,195
44.20
11,868,228
Venezuela,
RB
31,358
136,686
25,275,281
32.00
8,088,090
Zimbabwe
3,103
14,098
12,746,990
64.20
8,183,568
MIC Totals
418,112
11,360,906
4,338,293,207
2,058,063,861
World
Totals
806,722
23,895,742
6,482,257,297
2,706,036,650
% of World
Total
51.83
47.54
66.93
76.05
Countries are those eligible to borrow from the IBRD as of
December, 2005.
2
Source: World Resources Institute EarthTrends
(http://earthtrends.wri.org/).
3
Source: United Nations World Population Prospects Database
(http://esa.un.org/unpp/).
4
Source: World Bank/WDI, supplemented by PovCalNet.
1
CGD0502 0527_Engl_6x9.indd 149
8/17/06 2:55:27 PM
150
Rescuing the World Bank
Notes
1. The author is grateful for the invaluable contribution
of Anthony Ody to the overall preparation of this chapter,
and for research assistance from William Gee.
2. The term “middle-income countries” refers here
to those eligible to borrow from the World Bank’s nonconcessional IBRD (International Bank for Reconstruction
and Development) window, which lends at interest rates
slightly above the World Bank’s own cost of borrowing in
the international capital markets. By contrast, “low income
countries” mostly borrow from the Bank’s concessional
IDA (International Development Association) window at
substantially softer terms, with the flows funded largely
from periodic “replenishments” voted by the Bank’s
more affluent shareholder countries (supplemented by
internal transfers from IBRD earnings). A few countries
borrow simultaneously from IDA and IBRD: these “blend”
countries are for most purposes counted within the
“middle income” classification.
3. Allan H. Meltzer, chairman, Report of the International
Financial Institutions Advisory Commission (Washington,
D.C., 2000), available at http://www.house.gov/jec/imf/
ifiac.htm.
4. “Charting a Way Ahead: the Results Agenda”
Address to the 2005 Annual Meetings by Paul Wolfowitz.
September 24, 2005. World Bank.
5. World Bank News Release No. 2006/205/S
(December 13, 2005).
6. The strategic importance of the middle income
countries for the realization of many international goals—
and for donor countries supporting these goals—are
addressed in greater detail in a later section of the
chapter.
7. Data for fiscal year 2006, released just
before publication of this volume, confirmed the
continued recovery of IBRD approvals—up another
4 percent above fiscal 2005, to the highest level
in seven years ($14.1 billion). Data available at:
http://web.worldbank.org/WBSITE/EXTERNAL/NEWS/
0,,contentMDK:21016240~pagePK:64257043~piPK:
437376~theSitePK:4607,00.html.
CGD0502 0527_Engl_6x9.indd 150
8/17/06 2:55:27 PM
151
Selected Essays
8. See especially The Role of the Multilateral
Development Banks in Emerging Market Economies,
the report of a commission co-chaired by José Angel
Gurria and Paul Volcker (Washington, D. C.: Carnegie
Endowment for International Peace, 2001), and The
Hardest Job in the World: Five Crucial Tasks for the New
President of the World Bank, report of a Center for Global
Development working group co-chaired by Nancy Birdsall
and Devesh Kapur, in this volume.
9. See The Hardest Job in the World, Birdsall and
Kapur (2006) for more.
10. The World Bank Annual Report 2005, (Washington,
D. C.: The World Bank, 2005).
11. Meltzer, Report of the International Financial
Institutions Advisory Commission.
12. Note, too, that while the commission’s text refers
only to cutting off countries with “investment grade,”
some of Prof. Lerrick’s comments in the present debate
implicitly question the rationale for support even to
countries with below-investment grade ratings (more
commonly known as “junk” ratings).
CGD0502 0527_Engl_6x9.indd 151
8/17/06 2:55:27 PM
CGD0502 0527_Engl_6x9.indd 152
8/17/06 2:55:27 PM
The Missing Mandate:
Global Public Goods
by Michael Kremer
T
his note discusses the potential role of the World
Bank in providing global public goods. From an
economic point of view, global public goods are
those for which a large share of the benefit cannot be
contained within a single country. For instance, a country
that establishes a policy to reduce carbon emissions
to prevent global warming does not just benefit itself
but helps all countries that would be hurt by global
warming. Likewise, a successful campaign to eradicate
polio would not only improve health in those countries
where the disease persists, but would also save other
governments hundreds of millions of dollars a year in
avoided immunization costs.
It’s worth noting that there is a continuum in the extent
to which the benefits of goods spill over across borders.
Reductions in carbon emissions or efforts to eradicate
polio are at one end of the continuum, while efforts to,
for example, expand the use of nets to fight malaria are
at the other, since most epidemiological models would
suggest that the great majority of the benefits of net
programs fall within a country, although theoretically there
might be some reduction in transmission of disease to
the neighboring countries. (As this example illustrates,
not every worthwhile investment is a global public good,
and the international community should not feel that every
activity undertaken by international organizations needs
to be justified as falling under this rubric.)
The full value of global public goods is not reflected
in an individual country’s own cost-benefit analysis. As
such, governments often have inadequate incentives to
devote their own resources to these programs.
Michael Kremer, Harvard University and
Center for Global Development
Michael Kremer is professor of economics at Harvard University, senior
fellow at the Brookings Institution, and non-resident fellow at the Center for
Global Development. Kremer serves as associate editor of the Journal of
Development Economics and the Quarterly Journal of Economics. He is an
expert on AIDS and infectious diseases in developing countries, economics
of developing countries, education and development, and mechanisms for
encouraging research and development.
CGD0502 0527_Engl_6x9.indd 153
153
8/17/06 2:55:27 PM
154
Rescuing the World Bank
International organizations such as the World Bank fill
this void by supporting global public goods. Of course
some global public goods, such as development of
improved algorithms for matching kidney donors to
patients, would benefit primarily rich countries. For equity
reasons, the World Bank may want to focus on those
global public goods for which a large share of the benefit
goes to poor countries.
When the Bank was limited to providing loans to
governments, the Bank’s instruments for promoting global
public goods were likewise restricted. It is difficult for the
World Bank to ask a country to repay any substantial part
of a loan if the benefits are primarily for other countries.
Because it now can offer grants, the Bank has the potential
to do much more to promote global public goods that
benefit the poor. It may make sense for the Bank to set
up a funding mechanism for these global public goods
that is separate from IDA.
In particular, the Bank might consider a separate
funding mechanism that could make investments in the
following global public goods: 1) technologies for the poor,
2) developing knowledge about what works in public policy,
3) a road network for Africa, and 4) creating incentives
for countries to house and care for refugees.
Technologies for the Poor
The development of certain health and agricultural
technologies, such as a schistosomiasis vaccine or
improved cassava varieties, would affect the lives of people
in many developing countries. While private companies
are often motivated to develop new technologies based
on a combination of up-front public funding as well as
the prospect of a market, markets for these technologies
either do not exist or function poorly.
The Bank can help overcome these hurdles in two
ways. First, it can use its funds to support research. In
the past the Bank has used part of its profits to support
the development of agricultural technologies through
institutions like the CGIAR system. The Bank could expand
its support of these activities and of the development of
health technologies as well.
The second approach would be for the Bank to use
its resources to create a market for the needed products
and thus create incentives for private firms to invest in
these technologies. In particular, the Bank could make a
commitment to extend loans or grants to help countries
CGD0502 0527_Engl_6x9.indd 154
8/17/06 2:55:28 PM
155
Selected Essays
finance the purchase of certain goods, like the purchase
of a schistosomiasis vaccine.1
Moreover, if the Bank had a separate financing
mechanism devoted specifically to supporting global
public goods, it could do much more than lend money
to countries to buy vaccines. The Bank could also offer
up-front to devote its resources to the creation of new
vaccines, through an Advance Market commitment
for example.
Developing Knowledge on Public Policy
A second global public good that targets poor countries
is the development of a solid knowledge base on the
impact of various public policies in these settings.
Oportunidades, the conditional cash transfer program in
Mexico formerly known as PROGRESA, is a prime example
of the benefits of combining policy innovation with rigorous
evaluation to determine their impact. Mexico developed a
very sophisticated evaluation mechanism for Oportunidades,
which included randomizing the order in which the program
was phased in across villages. The high quality evaluation
techniques created a reliable evidence base that has not only
helped lead Mexico to preserve and expand the program,
but also led other countries to adopt similar programs. (The
IDB played a role in this, and the Bank has played a role in
expanding this type of program.)
The Bank could support countries interested in testing
new approaches through an innovation and evaluation
fund. Through such a fund, the Bank could make resources
available, with a large grant element, to countries that are
willing to subject programs of potential interest to other
countries to a rigorous evaluation, including randomization
of the order of phase-in.
African Road Network
Support for an African road network would provide a
regional public good to one of the poorest areas in the
world. (Since there are other financial institutions that
specifically serve this region, some have argued the Bank
follow the subsidiarity principle by funding or offering
co-financing to those institutions, such as the African
Development Bank.)
If the Bank were to consider such a program, it would
be critical to provide for a continued role of international
institutions in preservation and maintenance of the road,
including preventing the overloading of trucks, which
damages roads. In the absence of such an international
CGD0502 0527_Engl_6x9.indd 155
8/17/06 2:55:28 PM
156
Rescuing the World Bank
role, there may not be sufficient incentive to prevent
road damage from heavy trucks since much of the cost
of that kind of road deterioration would be shared with
neighboring countries.2
Support for Refugees
More speculatively, support for refugees can be considered
an international public good. Each country would rather
that another accept refugees. This is one justification
for the international treaties which require signatories to
accept political refugees. Yet the current system creates
a number of problems.
Maintaining refugees in camps concentrates them in a
particular area, making it harder for them to work in the
normal economy, and leaving them plenty of time and
incentive to engage in politics, including violent political
activities. We have seen this most recently with Rwandan
refugees in Congo. We also saw it in Afghanistan, and
with the Palestinians.
In some instances, powerful political actors may have
political motivations for keeping refugees in camps. For
instance, some may have felt that in order to keep pressure
on Israel, it was strategically useful to have Palestinians
in refugee camps rather than dispersed and resettled
throughout the Arab world. On the other hand, there
are certainly some cases—the Hutu militias and other
refugees that fled to Congo might be a good example—in
which there is no strong international political force which
is lobbying for refugees to be maintained in camps, and
there might be prospects for reform.
One way the Bank might assist in such situations would
be to support countries, especially non-neighboring
countries, that are willing to take in refugees and to let
them integrate into their society. For example, if the World
Bank had a mandate in this area, it could have provided
assistance to countries that were willing to take in
refugees on condition that they would allow them to settle
freely. For example, there could have been assistance to
countries like Kenya if they accepted Rwandan refugees.
Had this occurred, it is possible the invasion of Congo
and the terrible war there could have been avoided.
CGD0502 0527_Engl_6x9.indd 156
8/17/06 2:55:28 PM
157
Selected Essays
Conclusion
Global public goods present a problem to the international
community. Because the returns on an investment in these
goods are shared around the world, individual countries
rarely have the incentive to devote their own resources to
providing them. The World Bank, armed now with grant
instruments in addition to loans, is in a unique position
to support the creation and maintenance of these goods.
It would be appropriate for the Bank to focus on the
Notes
1. See Michael Kremer and Rachel Glennerster,
Strong Medicine: Creating Incentives for Pharmaceutical
Research on Neglected Diseases, (Princeton, New
Jersey: Princeton University Press, 2004); Owen Barder,
Michael Kremer and Ruth Levine, Making Markets for
Vaccines: Ideas to Action, (Washington D.C.: Center for
Global Development, 2005); Michael Kremer and Rachel
Glennerster, “A World Bank Vaccine Commitment,”
Brookings Institution Policy Brief #57 (May 2000),
http://www.brookings.edu/comm/policybriefs/pb57.htm.
2. See Nancy Birdsall, “Underfunded Regionalism
in the Developing World,” in The New Public Finance:
Responding to Global Challenges, eds. Inge Kaul and Pedro
Conceição, (Oxford: Oxford University Press, 2006).
CGD0502 0527_Engl_6x9.indd 157
8/17/06 2:55:28 PM
CGD0502 0527_Engl_6x9.indd 158
8/17/06 2:55:28 PM
The “Knowledge” Bank
by Devesh Kapur
I
deas have always been a core trait of the Bank. Indeed,
if the Bank was simply a financial intermediary, it would
barely need a tenth of its staff.1 The money was seen
as the lubricant to move the main product—ideas on what
to do, how to do it, who should do it and for whom. In
its early decades, this importance of ideas was implicit.
The source of ideas was the knowledge embedded in its
experienced personnel and the transmission mechanism
was the project mode. Since the 1970s, however, the
Bank became more self-conscious about the importance
of knowledge, both as an imprimatur institution as well as
a producer of knowledge. And in more recent years, the
relative decline in the importance of the Bank’s financial
role (especially in emerging markets), in part the result of
the high transactions costs of borrowing from the Bank,
has led to a greater stress on its role as a “knowledge”
intermediary rather than just (or even primarily) as a
financial intermediary. However, lending has been the
principal mechanism for knowledge transfer, and any
stagnation or decline in lending is likely to adversely
impact knowledge transfer as well.
The World Bank’s extensive (and expensive)
commitment to the production and dissemination of
knowledge has led to substantial critical analysis of the
Bank’s “knowledge” activities. Yet there is an analytical
vacuum on key issues that bear on the subject, be it the
optimal quantum of budgetary resources allocated to this
area, the distribution of those resources among different
research activities, between generation and diffusion, or
the optimal institutional mechanisms to generate and
transmit the research, whether in-house or externally.
Devesh Kapur, University of Pennsylvania and
the Center for Global Development
Devesh Kapur holds the Madan Lal Sobti Professorship for the Study
of Contemporary India and is the Director of the Center for the Advanced
Study of India at the University of Pennsylvania. He is also a non-resident
fellow at the Center for Global Development. His recent publications include
Give Us Your Best and Brightest: The Global Hunt for Talent and Its Impact
on the Developing World (Center for Global Development, 2005) and Public
Institutions in India: Performance and Design (Oxford University Press, 2005).
CGD0502 0527_Engl_6x9.indd 159
159
8/17/06 2:55:28 PM
160
Rescuing the World Bank
Consider for instance the following hypothetical
questions:
1. If the Bank’s overall budget was cut by a third and
the resulting savings (more than half billion annually)
were put into research in those diseases, crops and
energy technologies that are sui generis to poor
countries, would the global welfare of the poor improve
or decline?
2. If the Bank were to cut its “Analytical and Advisory
Activities” (AAA) expenditures (from its estimated $600
million in 2005), shifting its focus from the social sciences
to funding research in the health sciences, would the
global welfare of the poor increase or decline?
3. If the Bank’s research activities were more akin to
a National Science Foundation (NSF) type funding
activity, rather than in-house research, would LDCs
gain or lose?
4. If the Bank were to reallocate part of its large transfers
from net income (about $600 million annually over
the past few years) to create endowments for
centers of learning in LDCs, would those countries
be better off?
This paper argues that the World Bank should give
greater emphasis to financing rather than producing
research, in particular, financing developing country
research institutions. Despite the modest quality of the
latter, such a shift is likely to be more effective in changing
national policies and in nurturing implementation. It will
also contribute to long-run institution building (at a
minimum, by not reinforcing the brain drain).
Although a large array of studies has demonstrated
the high rates of social return in publicly funded research,
this in itself does not provide any guidance either on
which areas to finance investment in, nor the precise
mechanism to undertake this task.2 High average values
for publicly funded research are of little use in deciding
whether to increase (or decrease) funding for public
research, or in choosing the mechanism that would yield
the best results (resource allocation decisions require some
sense of marginal rather than average rates of return).
Moreover, there is no analytical framework that would help
answer whether the World Bank should conduct research
CGD0502 0527_Engl_6x9.indd 160
8/17/06 2:55:29 PM
161
Selected Essays
in-house, outsource it by funding universities or research
centers (and if so, create country—or sector-specific
centers), promote joint research ventures (including
exchange of personnel), or build research networks (such
as the Global Development Network).
The dilemmas are compounded by the reality that
research capabilities are located in the North while many
of the issue areas, with the highest rates of social return
to public investments in research, are in resource-poor
countries. Furthermore, even if the World Bank were to
outsource its research and fund more research, what
mechanisms should it follow? For instance, in areas
where research is undersupplied because of severe
market failures—such as tropical diseases, where
pharmaceutical firms do not invest fearing that were they
to actually develop a product, they would face severe
public pressure to sell the product at a price that would
not justify the initial investment—a novel mechanism
that has been proposed is for public agencies to finance
“tournaments” with a prize guaranteed to any entity that
meets predetermined specifications at a certain price.3
Although such an approach would not build developing
countries’ own capabilities, it might be warranted in areas
where delay has high human costs.
The stakes are different, however, in policy research,
the core focus of the Bank’s AAA. The background
of Bank researchers creates strong incentives to give
pride of place to propositional knowledge—the search
for “universal” laws of development from the frontiers
of academia—and using that to generate prescriptive
knowledge. LDC-based researchers are seldom, if ever,
represented in the former. Does that matter? There are
several good reasons why concerns on this score may
not be warranted. For one, there are typically participants
from developing countries in conferences focusing on
propositional knowledge. It just so happens that their
institutional base is in industrialized countries (typically
the United States). Second, the idea that one’s analytical
position is an isomorphic reflection of one’s nationality
and/or geographical base is rather specious. Third, one
could argue that the Bank should only be drawing on
the best talent to understand difficult issues, and if it
happens that the talent is based in North America, so be
it. Fourth, the fears of a lack of diversity are misplaced,
given the vigorous debates and differences that are
CGD0502 0527_Engl_6x9.indd 161
8/17/06 2:55:29 PM
162
Rescuing the World Bank
integral to academic and intellectual cultures. And finally,
the skewed participation may simply reflect the realities
of the global production of knowledge, in which LDCs
themselves have played a not insignificant role by running
their own universities and knowledge production systems
to the ground.
However, there are reasons for unease as well.
Intellectual networks can be double-edged. While
they reduce selection costs and serve as reputational
mechanisms, they can also be prone to a form of “crony
intellectualism.” This inherent tendency to inbreeding has
negative consequences for intellectual advancement.
Researchers, like other societal groups, also have
interests. And research involvement with the World
Bank has substantial payoffs, from research funding to
access to data and visibility. Moreover, the very nature of
academia means that academic researchers (in the social
sciences) are not accountable for the consequence in the
sense that their work responds to professional incentives,
not to its development payoffs. These professional
incentives place a large positive premium in academic
papers on the novelty of ideas, methodological innovation,
generalizability and parsimonious explanations. Detailed
country and sector knowledge, an acknowledgement that
the ideas may be sensible but not especially novel, that
uncertainty and complexity rather than parsimony are
perhaps the ground reality, are all poor country-cousins
of research that purports to find universal truths. The
mainline prestigious journals usually give short shrift
to articles with micro-data painstakingly collected in a
LDC. These journals act as gatekeepers of knowledge
as well as reputation but are important markers for the
Bank on the who, how and what dominates its research
agenda. For the most part, this service is positive, given
the concentration of talent in these institutions. But the
fact is that unless a researcher is part of this circuit, she
is marginalized.
This is also an important reason why the Bank’s
knowledge activities have underemphasized the crucial
long-term contributions of its didactic and educational
role. The very fact that the vast majority of the Bank’s work
on poverty is in English, a language understood by almost
none of world’s poor, indicates the low status assigned
to this role, and cannot be understood without reference
both to the internal incentives and external networks of
CGD0502 0527_Engl_6x9.indd 162
8/17/06 2:55:29 PM
163
Selected Essays
Bank staff which skew the priorities of research staff in
these institutions. The professional payoffs of delivering
a paper on Africa are substantially greater in Cambridge,
Massachusetts, than in Ouagadougou. In turn that means
that the questions and methodologies will be geared to
the priorities of the former, even though the latter has more
at stake. Growth regressions have undoubtedly helped in
the growth of researchers, but have they contributed to
the growth of poor countries themselves? The search for
universality offers little by way of prescriptive knowledge in
a particular situation. Yes, institutions matter, but anyone
examining the first few decades of the Bank would not
view this insight as a Eureka moment. In the end, such
prospective knowledge offers little insight into prescriptive
knowledge. Of the scores of institutions that matter, which
institution is most critical in which country at any specific
period of time cannot get around the need for its having
a deeply textured knowledge of the circumstances of the
country itself. And it offers even less by way of guidance
to the most glaring weaknesses of poor countries: how
to build these institutions and who would do so.
The virtual absence of researchers based in developing
countries in the more prestigious development
conferences cannot be attributed simply to exclusionary
networks. Given the outpouring of reports on key
global debates involving the World Bank, networks and
reputation are critical screening mechanisms. On both
counts, a base in a developing country virtually ensures
extinction. The developing countries—especially the
larger ones—have much to answer for themselves, having
failed to develop and maintain reputational institutions
in the social sciences.5 The poor quality of developing
country academic institutions in the social sciences
has led the Bank to not only draw its research staff
from U.S. universities in particular (which then creates
research networks between the staff and faculty in those
universities), but when these institutions want to train and
support developing country students or send their own
staff for training, it is invariably again at U.S. universities.6
Given the outstanding quality of the latter, the shortrun compulsions of the Bretton Woods institutions are
quite understandable, but their long-term consequences
are inimical. These practices have strengthened already
strong research institutions in the U.S. while further
weakening developing country institutions—creating
CGD0502 0527_Engl_6x9.indd 163
8/17/06 2:55:29 PM
164
Rescuing the World Bank
conditions for perpetuating the practice. The process
has generated a vicious circle with results that are in line
with models of statistical discrimination. The more the
World Bank and the IMF in effect discriminate against
researchers from LDCs, the more the incentive of these
researchers to migrate out of the countries either to these
institutions themselves or to developed countries where
their credibility is enhanced by their association with a
developed country institution, furthering the decline of
LDC research institutions.
It is not that there are no universal “truths” about
development, but rather that they make the Bank a
prisoner as often as they liberate the institution from
past mistakes. Consequently, the Bank’s knowledge
activities have been more captive to the fads and fashions
of academia, moving from one big idea to the next, rather
than knowledge activities that might be most helpful to
its borrowers. Fig. 1 is a schematic representation of
these cycles, where new ideas lead to new projects and
programs, with recruitment and expertise usually lagging.
As time passes, evaluations and feedback usually paint
a more somber picture, requiring course correction. But
even as the knowledge resulting from learning-by-doing
begins to get accumulated, a form of intellectual ennui
sets in and a new set of ideas (often precipitated by a
change in guard at the top), begins a new cycle. From
rural development in the 1970s to structural adjustment
in the 1980s to institutional changes such as judicial
reforms in the 1990s, much new knowledge has been
learnt in the Bank—and forgotten as it is crowded out
by new ideas and agendas.
Consequently, the substantial resources devoted by
the Bank to self-evaluation have had limited effects—the
sum being considerably less than its parts. The evaluation
methodology has been questioned, in particular, on
whether in the absence of randomized trials, lessons
from these evaluations can be meaningfully extrapolated.
While a valid criticism, the growing fashion for randomized
trials glosses over the reality that while providing valuable
insights for a particular context, they too have weaknesses
in the lessons they provide for similar projects but in
different contexts. The more troubling weaknesses are
that the tacit knowledge born of experience has become
a premium in the institution as the average experience of
Bank staff has fallen sharply. While new blood is always
CGD0502 0527_Engl_6x9.indd 164
8/17/06 2:55:29 PM
165
Selected Essays
critical, periodic reorganizations and perennial transfers
have ensured that loud displays of innovation are often
old wine (if not old vinegar) in new bottles, born more
of inexperience than perspicacity. But in the end, even
the best of evaluation techniques and self-knowledge
are at the mercy of the willingness and ability of the
Bank’s top echelons to be open-minded and guided by
empirical knowledge. That in turn is a function of the
Bank’s governance.
As a result, a half century into “development,”
developing countries still seem incapable of thinking for
themselves on issues (to put it crudely) critical to their own
welfare, at least as measured by the lack of meaningful
contributions that would find a place at the high seats
of social science research. What has the Bank done in
the last half-century to build institutions in developing
countries that could help them think for themselves?
For the most part, the answer is “not much.” Even
as MNCs increasingly have diversified the geographical
location of research activities, research is still relatively
centralized in the Bretton Woods institutions—and to
the extent that ideas shape agendas, centralized control
of research is an excellent unobtrusive approach to set
the agenda. Large salary differentials offered by these
institutions and developing country research institutions
Fig.1. The Life Cycle of Ideas and Output
in the World Bank
Volume
Ideas
Projects
Expertise
New cycle of Ideas
Time
Project Quality
Feedback
CGD0502 0527_Engl_6x9.indd 165
New Ideas / New President
8/17/06 2:55:30 PM
166
Rescuing the World Bank
(with the exception of some Latin American countries)
means that they often draw out limited talent in developing
countries. Moreover, for nearly two decades the Bank has
been chary of supporting institutions of higher learning,
directing resources to primary and secondary education
and justifying this shift both on equity and efficiency
grounds. Foundations have also joined the bandwagon
against supporting research institutions in developing
countries on the grounds that they were elitist and that
instead, “grass-roots” institutions needed more support.
In both cases there was more than ample justification for
the shift—but in the process, Bank (and the Foundations)
have thrown the baby out with the bathwater. It has
meant that developing country researchers are by and
large restricted to data collection and country-specific
applied work, not only incapable of contributing anything
meaningful to agenda setting debates ranging from global
financial architecture to second generation reforms, but
remaining dependent on continuing and often secondrate technical assistance that is also very expensive.
Should the Bank move from a producer to a
financier of knowledge?
As an intergovernmental organization, the World Bank’s
knowledge activities will always be subject to pressures
from members. If, in the 1980s, debt and corruption were a
no-mans land, in recent years intellectual property rights,
capital account liberalization and genetically modified
crops are examples of issue areas that the Bank has had
to tiptoe around. If the value of the Bank’s research as a
global public good is undermined by its perceived lack of
independence, other factors would appear to strengthen
the case of the Bank moving from a producer of research
to a financier of research.
First, there are substantial opportunity costs. It should
be emphasized that in-house knowledge-related activities
at the Bank are expensive, even when compared to
U.S. universities, let alone LDCs. Second, there are
important strategic benefits of publicly funded research
for developing countries, particularly the creation of
capabilities, through the vital links between research and
the supply of skilled graduates. To put it differently, the
process of research creates capabilities that allow for
better consumption or use of knowledge. Additionally,
public funding of research in different environments
CGD0502 0527_Engl_6x9.indd 166
8/17/06 2:55:31 PM
167
Selected Essays
plays an important role in the creation of diverse options.
The domination of a narrow set of institutions (reflecting
in part their outstanding quality) has several undesirable
consequences. It skews the questions, methodologies
and other priorities of research. As a result, those
directly affected by the policies of the institutions are
underrepresented in setting the research and policy
agenda. Furthermore, it narrows the diversity of views,
which, given limited knowledge and the possibility of wrong
advice, could amplify risk in the international system. The
importance of diversity is particularly important in the
context of an uncertain future.7 Moreover, diversity may
matter in and of itself on the grounds that there should
be at least a minimum degree of participation by those
likely to be affected by the actions resulting from ideas
emanating from these institutions. Diversity may also be
important for its instrumentality—it diversifies risk, a not
unimportant criterion, given limited knowledge and the
consequences of misplaced advice.
The rhetoric of the Word Bank and IMF on institutions
notwithstanding, they have been tepid in supporting
initiatives to develop knowledge-producing institutional
capacity in LDCs, although over the last decade the
World Bank has made some efforts to support regional
research centers.8 Its support for the Global Development
Network (GDN, which has now been spun off as an
independent entity) is an interesting innovation aimed
at linking researchers and policy institutes involved in
the field of development. The network also aims at skill
and reputation building. This is a commendable effort,
although it is too early to gauge its impact. However,
even the GDN is unlikely to address the problem of how
developing country researchers can overcome the high
reputational barriers that exist on research and policies
related to systemic issues. That requires a receptivity
and openness in these institutions themselves, which is
structurally difficult. Virtually all the links in the research
Web sites of the Bank and Fund are to researchers in
developed countries, a reflection of the modest quality of
research from LDCs but also an indication of the personal
networks of research staff in these institutions.
Consequently, it would appear that all factors, from
operating costs to opportunity costs (using the resources
to build capabilities in LDCs), would seem to support a
serious reconsideration of the allocation of AAA-related
CGD0502 0527_Engl_6x9.indd 167
8/17/06 2:55:31 PM
168
Rescuing the World Bank
resources by the World Bank. But not surprisingly, there are
countervailing factors as well. First, conducting in-house
research has operational externalities for the Bank. The
possibility of being able to undertake research at the Bank
attracts higher quality personnel (especially economists)
who then contribute positively to the operations side of
the Bank. Moreover, what is true of developing countries
is also true of the Bank: in-house research capabilities
increase the ability to sift through the copious volumes
of new knowledge and ideas and make better judgments
in separating the wheat from the chaff. Second, and
contrary to popular impressions, it is easier for the Bank
to restructure its internal AAA, than restructure its support
to external research institutions.9
But even if the World Bank were to finance knowledge
activities in LDCs to a greater degree, a different set of
dilemmas arises—should the activities be focused on
knowledge activities that are more national or global?
While the case for the latter seems evident, in some issue
areas the quest for supplying knowledge-related public
goods at the global level may be amplifying the deficit
at the national level. Agricultural research is a case in
point. According to one estimate, even in the 1980s, while
nearly a third of the hundreds of agricultural researchers
who routinely attended the CGIAR’s annual “Centers’
Week” meetings at the World Bank were originally from
LDCs, more recently only about one in 20 were still
actually affiliated with LDC national research institutes
or universities. With donors viewing the building of higher
education and research capacity in LDCs as “elitist,”
research as a public good is seen to be better supplied at
the global rather than the national level. However, it may
well be the case that in areas ranging from agricultural to
economics research, LDC researchers faced with rewards
that are much greater in international rather than national
research organizations, gravitate toward the former. As a
result, while the supply of global public goods (in the form
of research in agriculture and economics) is reasonably
adequate, public goods deficits at the national level,
involving the production of country-specific knowledge,
may be increasing.
Conclusion
There is no development institution that has devoted as
many resources to knowledge-related activities as the
CGD0502 0527_Engl_6x9.indd 168
8/17/06 2:55:31 PM
169
Selected Essays
World Bank. It is therefore surprising that the Bank has
had little appetite to develop a rigorous framework that
at a minimum analyzes the opportunity costs of these
substantial knowledge-related expenditures. Admittedly
the task would be analytically difficult, but there are few
incentives within the institution to do so. Arguably, if even
a tenth of this expenditure was instead redirected at
creating endowments for knowledge-producing centers in
developing countries, it is at least an open question if the
welfare of those societies may not be higher. It may help
LDCs to think for themselves—and take responsibility for
the actions resulting from their ideas—rather than be the
perennial objects of received wisdom.
Notes
1. As is the case with the European Investment Bank,
whose loan portfolio has been larger than the Bank but
which is otherwise a much smaller institution.
2. See, for instance, A.J. Salter and B.R. Martin, “The
economic benefits of publicly funded research: a critical
review.” Research Policy 30 (2001): 509–532.
3. Of course the prize must be large enough to be
attractive for organizations to undertake the investments.
For an application to vaccines, see Michael Kremer,
“Creating Markets for New Vaccines—Part I: Rationale,
Part II: Design Issues,” Innovation Policy and the Economy,
Vol. 1 (2000): 35–118.
4. Pranab Bardhan, “Journal Publication in Economics:
A View from the Periphery,” The Economic Journal 113
(2003): F332-F337.
5. The case of India is illustrative. In the 1990s, of
the 2,312 articles in the top five journals in economics
(AER, EJ, JPE, QJE and Review of Economic Studies),
138 were by Indians outside India and just seven were
from Indians in India—a factor of 20!
6. At the beginning of the 1990s, 80 percent of the
research staff at the World Bank had graduate degrees
from U.S. and U.K. institutions (nearly two-thirds from
the U.S.). While similar data from the IMF is unavailable,
CGD0502 0527_Engl_6x9.indd 169
8/17/06 2:55:31 PM
170
Rescuing the World Bank
it is unlikely to be less. Since then, widening quality
differences between U.S. and developing country
academic institutions are likely to have increased
the skewedness. See Nicholas Stern, “The World Bank
as Intellectual Actor,” Table 18-12-6, in Devesh Kapur,
John Lewis and Richard Webb, The World Bank: Its
First Half Century (Washington, D.C.: The Brookings
Institution, 1997).
7. Andrew Stirling, “On the Economics and Analysis of
Diversity.” SPRU Working Papers, University of Sussex,
Brighton (1998).
8. These include the Africa Economic Research
Consortium (AERC) and the Joint Vienna Institute
(cosponsored with the BIS, the EBRD, the IMF and
the OECD). But the output of these institutions is not
geared to addressing systemic issues—as attested by
the fact that their work is rarely cited by the sponsoring
institutions themselves. The IDB has been more creative
in this regard. It has been coordinating the Latin American
Research Network, created in 1991, and funds leading
research centers in the region to conduct research on
economic and social issues in Latin America and the
Caribbean. The research topics are determined through
consultation with IDB and external professionals.
9. This is exemplified by the Bank’s large financial
support over more than three decades to the CGIAR
system. There can be little doubt that certain centers need
to be closed down and the limited resources shifted to
others, but for political economy reasons this has been
stymied, leading to a sub-optimal allocation of resources
within the CGIAR system.
CGD0502 0527_Engl_6x9.indd 170
8/17/06 2:55:31 PM
Evaluations and Aid Effectiveness
by Pierre Jacquet1
A
ll bilateral and multilateral development institutions
devote important resources to evaluating their
operations. This activity is generally assigned
to a dedicated unit placed outside the purview of the
operational sphere so as to protect it from any vested
interests; it is also often conducted in partnership with
independent experts. Although they account for only
a tiny proportion of public budgets, bilateral official
development aid institutions probably even stand
several steps ahead of many public administrations in
thus contributing to the evaluation of public policies.2
This brief paper, based on the experience of a bilateral
development agency, claims that evaluations in fact fulfill
several distinct, albeit related, valuable functions and that
it is worthwhile to address each of them for their own
sake as methods and organization need to be tailored
accordingly. These are layman observations, however,
that do not aim at constituting a “theory” of evaluations,
but simply at providing some perspectives about an old
but increasingly important function or set of functions in
development institutions.
This paper does not primarily address the working of
the World Bank Independent Evaluations Group (IEG,
formerly OED—Operations Evaluation Department).
I believe, however, that bilateral and multilateral
development institutions all face similar challenges and
should exchange more on their evaluation approaches
and practices in order to improve on these practices and
set common standards through a careful benchmarking
process. Indeed, the description of IEG’s evaluation tools
and approaches3 cover much of what I describe below
as full-fledged “evaluation”, even though I propose here
Pierre Jacquet, Agence Française de Développement
Pierre Jacquet has been executive director (in charge of strategy) and chief
economist at Agence Française de Développement (the French Development
Agency) since 2002. He was formerly deputy director of the French Institute
of International Relations in Paris and chief editor of its quarterly review
Politique Etrangère. He is professor of international economics and chairman
of the Department of Economics and Social Sciences at Ecole nationale
des ponts et chaussées. Between 1997 and 2006 he was a member of the
Conseil d’Analyse Economique, an independent advisory panel created by
the French Prime Minister.
CGD0502 0527_Engl_6x9.indd 171
171
8/17/06 2:55:31 PM
172
Rescuing the World Bank
a much more explicit separation between the various
functions that I identify. In analyzing these functions, I
also try to highlight some organizational consequences.
Not everything needs to be “independent” or conducted
outside the operational sphere. And, insofar as scientific
impact evaluation is concerned, there is a powerful case
for some form of coordination among development
institutions.4
1. Current evaluations: a multi-purpose activity
While there is a welcome attention currently given to the
development of scientific impact evaluations geared to
seriously assessing the actual impact of ODA financed
operations on development, two related observations
stand out from the experience of development institutions:
first, not much has been done in terms of conducting
genuine scientific development impact measurements
so far; second, this is in fact not what most development
institutions call “evaluation” in the first place. Yet, existing
evaluation units play important roles that it is worthwhile to
recognize and strengthen. They typically simultaneously
fulfill three useful, different, functions, while admittedly
not serving other, useful purposes that do belong to
evaluation and that are in short supply in the international
community of donors.
Building knowledge on processes
and institutions
The first role is a cognitive role of building operational
knowledge of processes and practices within a
development institution and about how they interact
with practices, behaviors and institutions on the receiving
side. Additionally, evaluations build knowledge about the
countries in which donors operate, their societies, their
institutions, their needs, the quality of their governance. In
short, they contribute to an empirical field knowledge that
helps staff build field experience and become seasoned
to the intricacies and complexity of development aid. This
aspect of “evaluations” is important for the institution
and needs to be conducted in-house because it is part
of a necessary process of self-knowledge and selfeducation. It is also key to improving practices. Critiques
comparing that process with scientific impact evaluation
are based on a confusion of roles. What such evaluations
do amounts to capitalizing on the existing experience
CGD0502 0527_Engl_6x9.indd 172
8/17/06 2:55:32 PM
173
Selected Essays
and creating transmissible knowledge from experience.
They inform on processes and on the interaction between
donor processes, practices and behaviors and institutions
on the receiving side. They also point to shortcomings
that can be remedied in subsequent operations. They
can also be complementary to impact evaluations to the
extent that they inform on why observed results may
have been achieved.
A key aspect of evaluations, insofar as this first role is
concerned, relates to the feedback process through which
the results from these evaluations, however conducted,
will inspire new actions. Substantial improvement needs
to take place on feedback and retroaction, even though
these dimensions have been recognized for a long time
and have often led to substantial quality improvement.
How to make feedback more systematic should be a
top management priority. There is a natural tendency
to focus on the “production of knowledge” aspect of
evaluations rather than on how to use that knowledge
in daily operations. Insufficient time and resources are
devoted to that process and the structure of incentives is
more favorable to conducting evaluations than to learning
from them in the operational process. Indeed, a key to
“results-based management” should not be as much
about designing success, a rather elusive goal, as about
learning from failures. This is certainly an area in which
much progress needs to be achieved. A former United
Nations Food and Agriculture Organization (FAO) officer
for example described in his book5 how it took ten years
for the FAO to abandon its recommendation that farmers
use centralized storage facilities and villages establish
cereal banks despite a record from evaluations that such
centralization was inefficient and not viable.
While it is important, as we argue below, to add other
dimensions to evaluations, it is also crucial to devote more
time and resources to the feedback loop through which
knowledge helps improve overall quality. This does not
apply only to knowledge built through evaluations. An
important aspect of quality management hinges on the
ability of any institution to use all relevant knowledge in
shaping its actions. Good practices must therefore ensure
that decision processes do take into account results from
past evaluations, but also current knowledge existing
outside the institution and that needs to be identified
and collected. Informing decision processes through
CGD0502 0527_Engl_6x9.indd 173
8/17/06 2:55:32 PM
174
Rescuing the World Bank
these various channels is a priority.6 A major difficulty,
however, hinges on what is, or is not, relevant when taking
a decision on whether to finance an operation, given
the differences between past and current operations in
terms of overall political, institutional, economic, social
and technical environment.
Monitoring quality
Evaluations also fulfill an important monitoring role, that
can be part and parcel of the practice of evaluations
described in the preceding section, but that is of a different
conceptual nature. It amounts to checking observed
results against ex-ante expected ones and to monitoring
the execution of operations: What was the objective of the
program or project, was the money used for what it was
supposed to, did it reach the intended beneficiaries, what
was the time schedule for disbursements, how to explain
any difference between actual versus expected results,
and so on. This is an exercise in conformity and based on
identifying good operational practices and on monitoring
these practices. While the cognitive role alluded to above
is best achieved through retrospective evaluations, quality
needs to be monitored during execution and up to the
end of any given operation.
As such, monitoring needs to be conducted within the
operational sphere because this is where the relevant
information is to be collected. Regular auditing must
check that appropriate good practices and procedures
have been followed. Monitoring also supplies useful
information for the kind of evaluations that was mentioned
in the preceding section. Here also, feedback, as a way
to improve management, is an important dimension. It is
central to quality management and control. Monitoring is
thus an important instrument for top management.
External Accountability
As official development assistance deals with taxpayers’
money, organizing accountability is of course a central
task. There is, however, a necessary distinction to make
between accountability and judgment. Development
institutions are liable to provide an assessment of
what they do and to communicate it to the outside.
Accountability should amount to making all information
available to the outside and let outsiders judge from
that information and from their own perspective whether
institutions make a good use of public money or not.
CGD0502 0527_Engl_6x9.indd 174
8/17/06 2:55:32 PM
175
Selected Essays
This aspect of evaluation systems, however, is open
to challenge. There is an inherent conflict between the
internal and external uses of information produced by
evaluations. While the institution itself needs to learn from
its own mistakes, it has little incentive to communicate
outside on all potential mistakes even if it may gain in
credibility in being quite open and transparent. This is why
external evaluations are needed both to build institutional
credibility and to form a judgment on the quality of any
institution’s operations. They should be commissioned
and conducted in a fully independent way, outside
institutions (or at least organized by them in a mutual,
cooperative way so that peer pressure and peer review
allow to counter the inevitable bias of having an institution
judge its own actions). For example, the operations of
the French Global Environmental Fund (FGEF), a Fund
set up by the French government in parallel with the
French participation in the Global Environment Facility
(GEF) are thoroughly investigated before the decision
to replenish the Fund, every three years. A team of
independent auditors hired by the Board reviews the
accounts, strategies and operations of FGEF.
There is a further dimension of accountability and
judgment that could usefully be developed and that
the current evaluation systems have not taken on so
far, namely benchmarking of donors, multilateral as
well as bilateral. Such benchmarking would be useful
for two different sets of reasons. First, judgment of
donor operations lack appropriate benchmarks. Quality
of overseas development aid is better judged through
comparison than in absolute terms. Benchmarking will
better help inform taxpayers on whether their money
was well used and whether it served useful purposes
from their perspective. Second, benchmarking allows to
compare the relative performance of donors and creates
powerful incentives to improve efficiency in a world in
which there is increasing competition among donors to
collect development money, from a “market for aid” point
of view, not only between public donors, but also between
private foundations, NGOs, municipalities, etc.
May be even more than quality competition between
donors, however, information on which donor does
what best is also very valuable, as it may lead, within
certain limits set by the global objectives of bilateral and
multilateral assistance, to a possible natural division of
CGD0502 0527_Engl_6x9.indd 175
8/17/06 2:55:32 PM
176
Rescuing the World Bank
labor between donors. There are already some interesting
initiatives in that respect. The Consultative Group to Assist
the Poor (CGAP), a consortium of 31 public and private
development agencies that work toward expanding
access to financial services for the poor in developing
countries through micro-finance projects, took a very
interesting aid effectiveness initiative in 2002 in organizing
a peer-review of its member activities. Results were
discussed in several meetings and identified donors that
performed well and others that did not. Interestingly, at
least one donor found to perform poorly decided to retire
from microfinance operations as a result of this peer
review. Such an approach obviously could be replicated
in other areas.
2. Scientific impact evaluations
However well done and professional, current evaluations
are not up to the task of measuring the actual impact of
development projects and programs. Impact evaluation
is a demanding task, because it requires both careful
observation of direct and indirect results and careful
assessment of attribution of these results to the operation
that is under-evaluation. Control groups are needed,
and the whole process of impact evaluation should be
thought of very early on, as early as the operation to be
evaluated is itself identified. Recent progress in scientific
impact evaluation methods (be they random assignment
methods or other rigorous ones) now make it possible to
learn useful lessons about the actual development impact
of some operations, and there is a welcome move toward
developing this kind of approach. Impact evaluations are
surely not a panacea, but they give a renewed dimension
to evaluations, more akin to applied research than to the
ex-post assessment of operations traditionally undertaken
in evaluation units. It fulfills a fundamentally different role
of building scientific knowledge, and it is worthwhile to
encourage its development.
Decades of development assistance have brought
home the fact that development is a complex and
poorly understood process. This is a powerful reason
for focusing on empirical approaches that will highlight
the kind of actions and policies that achieve results and
those who do not and why. In turn, such knowledge is
necessary to allow for more effective selectivity within
development institutions, allowing them to focus on what
CGD0502 0527_Engl_6x9.indd 176
8/17/06 2:55:32 PM
177
Selected Essays
works and to avoid spending money on what does not.
More than an evaluation of development institutions
themselves, impact evaluation is a contribution to the
provision of a global public good, namely knowledge on
the development process.7
This public good character, together with the high
cost involved in conducting scientific impact evaluations
suggest that they will typically be under-supplied unless
the coordination problem of who does what is solved.
There is a further value added in coordination: better
knowledge about what works in terms of development will
have to be based not on empirical results from a single
operation, but from comparing results from a number of
operations in the same sector or with similar objectives.
The way forward is setting up a system of cooperation
between universities and development assistance donors
toward joint impact evaluation, in which many donors
take part, each of them contributing and all sharing the
knowledge that is produced.
Available scientific methods, such as random
evaluations, however, will work on some operations and
much less on others. For cost as well as practicality
reasons, not everything can undergo a scientific impact
evaluation. Not everything is amenable to, say, random
assignment like experimentation with drug use. For
the very credibility of any exercise in scientific impact
evaluation, therefore, it is crucial not to present it as a new
religion, but rather as a contribution to better knowledge.
For a start, what donors can usefully do is to initiate the
process by selecting a few appropriate projects on which
it is possible to build an operational cooperation between
academic specialists and operational project officers,
and by adopting a forthcoming, pragmatic approach. A
number of initiatives deserve to be encouraged, notably
the DIME project launched by the World Bank that
proceeds along such lines.8
As argued above, however, when discussing the use of
relevant knowledge in decision making, there will always
be a need to keep a critical eye on the results from such
scientific evaluations. It is unlikely that any scientific
method of evaluation will allow to grasp all relevant factors
in the interpretation of how a project or program fares in
a given context, especially the institutional, human and
societal dimension. What the method will help provide
CGD0502 0527_Engl_6x9.indd 177
8/17/06 2:55:33 PM
178
Rescuing the World Bank
is a scientifically informed knowledge base on actions,
not a book of recipes about development.
Moreover, it is useful to keep in mind that development
is more about processes than results. An objective of
“impact” evaluation should thus be to help measure
incremental improvement rather than final impact. Even
when positive impact is reached and documented, the
question remains about sustainability. From this point of
view, there is a continuum of concerns between process
and impact evaluations.9
3. Concluding Remarks
Part of the difficulty in debating the evaluation function
in donor institutions is that a number of different tasks
are implicitly simultaneously assigned to evaluation:
building knowledge on processes and situations in
receiving countries, promoting and monitoring quality,
informing judgment on performance, and, increasingly,
measuring actual impacts. Agencies still need their own
evaluation teams, as important knowledge providers
from their own perspective and as contributors to
quality management. But these teams provide little
insight into our actual impacts, and, although crucial,
their contribution to knowledge essentially focuses on a
better understanding of operational constraints and local
institutional and social contexts. All these dimensions of
evaluations are complementary. For effectiveness and
efficiency reasons, they should be carefully identified
and organized separately: some need to be conducted
in house, some outside in a cooperative, peer review
or independent manner. In short, evaluation units are
supposed to kill all these birds with one stone, while all
of them deserve specific approaches and methods.
There is a need to substantially buttress scientific
impact evaluations, because they clearly exhibit public
goods characteristics in terms of providing empirical
knowledge on development. They require increased
cooperation among donors and joint action. A number of
initiatives have been launched recently, notably under the
aegis of the World Bank. The focus on developing impact
evaluations, however, should not obfuscate the need to
considerably improve the operational feedback from
evaluation results and more broadly from all available,
relevant knowledge to operations.
CGD0502 0527_Engl_6x9.indd 178
8/17/06 2:55:33 PM
179
Selected Essays
Knowledge is not a scientific good only. Careful impact
evaluation is a complement, not a substitute, to the
non scientific, empirical approach that is also part of
knowledge building and quality control. As for judgment
of performance, this is clearly not a mission for the
donor agencies themselves: Their responsibility is to be
accountable, namely to provide all available, accurate and
unbiased information on their operations. Assessment
of performance needs to be totally externalized and
should not be even undertaken under a contract directly
commissioned by the donors themselves, except through
a carefully organized peer review system.
As a multilateral institution with a clear commitment
toward improving aid effectiveness and researching on
development processes, the World Bank should contribute
placing the role and format of evaluations higher on donors’
agendas. It has taken a leading role in developing impact
evaluations and engaging other donors in co-organizing a
publicly available knowledge base about the results from
such evaluations. In a recent report presented to Paul
Wolfowitz, President of the World Bank, in the spring of
2005, Birdsall and Kapur10 notably recommend that the
Bank should take the lead on independent evaluation of all
aid spending. It should first, however, be also exemplary
in taking part in existing peer reviews of donor operations
in specific areas. Regretfully, the Bank did not participate
in the CGAP peer review discussed above. Much is to
be said in favor of a joint action by donors, alongside the
CGAP example, to organize a rating of their operations
in specific areas. As CGAP has demonstrated, it can
provide very useful insights about what works and what
does not and help donors become more selective in their
operations.
Finally, development finance is not as much about
picking out operations that work as about taking informed
risks to discovering what works. Evaluations, along all
the dimensions discussed above, are most useful to
inform risk taking and decision making, rather than
to act as substitutes to risk. The central, operational
question is not whether operations similar to the one
under consideration have been demonstrated to work in
the past. In a dynamic, innovative, approach, it is rather
whether all relevant knowledge has been called and taken
into account so that the risk of development finance is
carefully assessed.
CGD0502 0527_Engl_6x9.indd 179
8/17/06 2:55:33 PM
180
Rescuing the World Bank
Notes
1. I thank Pierre Forestier, Sebastian Linnemayr,
Thomas Melonio, Jean-David Naudet and participants
in the Center for Global Development symposium for
useful comments, questions and suggestions. The usual
caveat applies.
2. In particular, it is puzzling—and of questionable
legitimacy—that we are seemingly putting more time
and energy in trying to properly organize the evaluation
of development projects than we seem to be in trying
to assess the effectiveness of public policies in our own
countries.
3. See www.worldbank.org/oed/.
4. For proposals along these lines, see William B.
Savedoff, Ruth Levine and Nancy Birdsall, “When Will
We Ever Learn? Recommendations to Improve Social
Development through Enhanced Impact Evaluation,”
Consultation Draft, Center for Global Development,
Washington D.C., September 15, 2005; and also Levine
and Savedoff in this volume. Through its Development
Impact Evaluation (DIME) project, the World Bank has
also undertaken an exercise about encouraging scientific
impact evaluations both in the Bank and in partner
development institutions, with the aim of collecting and
sharing results so as to improve knowledge on several
aspects of development processes.
5. Eberhardt Reusse, The Ills of Aid. An Analysis of
Third-World Development Policies, (Chicago: University
of Chicago Press, 2002).
6. This is a problem akin to the one studied in the
pioneering work by Richard Neustad and Ernest May,
Thinking in Time: The Uses of History for Decision Makers
(New York: The Free Press, 1986).
7. See for example Esther Duflo, “Evaluating the
Impact of Development Aid Programmes: The Role of
Randomised Evaluations,” in Development Aid: Why and
How? Towards strategies for effectiveness, Proceedings of
the AFD-EUDN 2004 Conference, Notes and Documents
No. 22 (Paris: French Development Agency, 2004).
8. The French Development Agency (AFD) has also
decided to invest in scientific impact evaluation, starting
with two projects, in microfinance and in health.
CGD0502 0527_Engl_6x9.indd 180
8/17/06 2:55:33 PM
181
Selected Essays
9. See the discussion in Smutylo, Terry (2001),
“Crouching Impact, Hidden Attribution: Overcoming
Threats to Learning in Development Programs”, Draft
Learning Methodology Paper, Block Island Workshop
on Across Portfolio Learning, 22–24 May 2001, Ottawa:
International development Research Centre.
10. Nancy Birdsall and Devesh Kapur, co-chairs, The
Hardest Job in the World. Five Crucial Tasks for the New
President of the World Bank, in this volume.
CGD0502 0527_Engl_6x9.indd 181
8/17/06 2:55:33 PM
CGD0502 0527_Engl_6x9.indd 182
8/17/06 2:55:33 PM
The Evaluation Agenda
by Ruth Levine and William D. Savedoff
The Bank’s Success Depends on Knowledge
W
e will start with an obvious point: To succeed as
an institution, the World Bank must succeed in
its main business. Its main business is financing
projects and programs that lead to better economic and
social conditions than would have occurred without those
projects or programs. A higher—and technically superior—
definition would require that the returns for these projects
and programs be at least a little better than their true
economic costs. And a still more demanding standard
might ask that the projects represent the best (most costeffective) of all possible ways to achieve the same ends.
But let’s not be fussy here; let’s just stick to the basic
message that the Bank succeeds when poor people’s
lives improve because of the funding, technical expertise,
accountability requirements or other dimensions of the
Bank’s lending and other instruments.
In contrast, the World Bank’s success cannot be
measured on the basis of whether the institution remains
solvent, gets along well with NGOs, keeps employees
happy, or fights corruption in-house and abroad. These
are all probably necessary, but they’re not sufficient. The
Bank’s success rests on whether it can make the lives
of those who are sometimes referred to as the “ultimate
beneficiaries” better off, in a meaningful way.
Ruth Levine, Center for Global Development
Ruth Levine is the director of programs and a senior fellow at the Center for
Global Development. She is a health economist with 16 years of experience
working on health and family planning financing issues in Latin America,
Eastern Africa, the Middle East, and South Asia. At CGD, she manages the
Global Health Policy Research Network. Before joining the Center, Ruth
designed, supervised, and evaluated health sector loans at the World Bank
and the Inter-American Development Bank.
William D. Savedoff, Social Insight
Bill Savedoff is a senior partner at Social Insight. In addition to preparing,
coordinating, and advising development projects in Latin America, Africa
and Asia, he has published books and articles on labor markets, health,
education, water, and housing. He currently serves on the editorial board
of Health Policy & Planning and as an editorial advisor to Transparency
International.
CGD0502 0527_Engl_6x9.indd 183
183
8/17/06 2:55:33 PM
184
Rescuing the World Bank
Whether the Bank’s projects and programs help
borrowing governments to achieve good results for
their citizenry depends in part on whether the programs
are designed and implemented well. Take, for example,
the case of a Bank-financed project aimed at improving
enrollment, retention and learning outcomes of primary
school students, which directs financing toward school
construction, curriculum development, teacher training
and new information systems in the Ministry of Education.
Whether the project will achieve the desired results
depends on whether the various project activities, from
contracting for civil works to developing manuals for
computer users, are conducted in a timely, cost-effective
manner that takes into consideration local conditions.
It also depends on whether the problem of low school
attendance and performance can be solved with
buildings, teacher skills, textbooks and computerized
enrollment records. One would imagine that guidance
on both of those questions, if not definitive answers,
are within reach. Given the Bank’s base of institutional
experience—more than 50 years, across more than
100 countries and every sector, with billions of dollars of
investments—a ready reserve of knowledge about what
works should be available to inform critical design and
implementation decisions. Indeed, it is just this type of
asset that inspired the notion that the Bank could be a
“knowledge bank.”
The reality is quite distant from this idealized
expectation, as any candid Bank employee will attest. The
Bank appears to be structurally and perennially unable
to learn.
The Bank Creates but Does Not Use
Operational Knowledge
For operational questions, the Bank has shown itself
to be reasonably good at generating “lessons”—but a
mediocre student when it comes to applying them. The
World Bank’s Implementation Completion Reports (ICR)
(prepared by staff or consultants at the conclusion of
every project) and the broader sector studies generated
by the Independent Evaluation Group (IEG), previously
called the Operations Evaluation Department (OED), are
replete with hard-won operational lessons, which are
conveyed to the Bank’s Board of Executive Directors in
confidential documents. Classic and oft-repeated ICR
CGD0502 0527_Engl_6x9.indd 184
8/17/06 2:55:34 PM
185
Selected Essays
conclusions include the inadvisability of establishing
project management units that are isolated from line
ministries; the importance of political commitment,
managerial continuity, and timely follow-through when
problems are detected; the need for operational research
to inform mid-course corrections; the benefits of focused
rather than multi-component “Christmas tree” investment
programs; and the importance of developing a realistic
financing strategy for the recurrent costs associated
with the program. Though the same mistakes may be
repeated from project to project, there’s no doubt they
are documented in detail each time.
The fact that these conclusions are oft-repeated is
testimony to the limited impact that their documentation
has on Bank practices and procedures, although they
are ritualistically invoked at particular moments. It is
striking, in fact, that one can often find essentially the
same “lessons” in both the design document justifying
the funding for a program and in the report after the
funds have been spent. The design document may say
that a “lesson learned” from similar operations is that
project activities should be clustered so that the newly
trained teachers are working in the rehabilitated schools
that have the additional textbooks. Then, the ICR for the
same project, seven years later, may say that results
were disappointing because the project had to disperse
investments widely, to maintain political support.
The reasons for this lack of learning about even the
operational basics are many, and include everything
from the extreme time pressure on staff, to the limited
funding for disseminating the ICR results in a meaningful
way, to the underlying incentives that result in oversized,
unwieldy, unrealistically ambitious projects. Essentially
no management attention is given to the sharing and
application of this knowledge; and ICRs tend to be seen
as bureaucratic by-products that yield no benefits to
line managers or those who design and supervise the
implementation of Bank projects.
The Bank Rarely Creates New Knowledge
about What Works
While the Bank at least documents the operational
lessons, it seldom generates the right kind of technical
knowledge, or knowledge about what really works to
achieve the desired impact. Technical lessons would come
CGD0502 0527_Engl_6x9.indd 185
8/17/06 2:55:34 PM
186
Rescuing the World Bank
from analyses of how well similar projects achieved their
aims in the past and would answer very basic questions
that are at the core of project designs: What are the most
effective (and cost-effective) ways to get girls to complete
secondary school in rural Africa? What AIDS prevention
strategies work to reduce the incidence of infection
among mobile populations? Under what conditions do
road-building projects reduce rural poverty?
On these sorts of questions, and the generation of
knowledge about what works, the Bank’s track record
has been as wanting as virtually all other development
institutions. It has systematically failed to even attempt to
learn from one project or program how to get more and
better results the next time around. Moreover, it has rarely
undertaken and shared the type of data collection and
analytic work that would contribute much needed light to
the darkness of development assistance more generally.
The type of knowledge needed comes from impact
evaluations, defined as evaluations that measure the
results of an intervention in terms of changes in key
variables (e.g., mortality, health status, school achievement
and labor force status) that can be credited to the to the
program itself, as distinguished from changes that are due
to other factors. That is, they are evaluations that permit
attribution of program-specific effects. At the Bank, as in
the field of development more broadly, much emphasis
has been placed on monitoring project performance and
comparing before- and after-project conditions, while
insufficient investments have been made in conducting
rigorous impact evaluations that are necessary to tell us
which interventions or approaches do and do not work.
This underinvestment in impact evaluation (and
consequent undersupply of evidence about the
relationship between specific types of investments and
their effects) has a major, if painfully obvious, result: If
we don’t learn whether a program works in changing the
well-being of beneficiaries, how do we know it’s worth
putting the money and effort into similar programs? If
we don’t bother to measure the results that are direct
consequences of the specific program, how can we make
a credible case for this, or any other type of expenditure
of public funds?
CGD0502 0527_Engl_6x9.indd 186
8/17/06 2:55:34 PM
187
Selected Essays
This is the current scenario, lacking accumulated
knowledge from impact evaluations. The typical World
Bank social sector project is designed with a narrow range
of inputs, sometimes generated by World Bank staff and
consultants, and sometimes by the government receiving
the loan or credit: a very detailed description and analysis
of current (bad) conditions; guesses about the reasons for
those bad conditions; a theory of change, or argument,
that says “if you make these particular sorts of investments
and institutional changes, the world will be a better place
in these specific ways: fewer children will get sick and
die, more children will go to primary school and learn
something that will permit them to make a living in the
future, and so forth. And not only will more of these good
things happen because of the program’s investments and
institutional changes, but those good things would not have
happened—or would not have happened so quickly—in
the absence of this program.”
Importantly, the dependence on this sort of argument
is central to even broad “country-driven” programs that
look much like budget support— for example, the Poverty
Reduction Strategy Credits (PRSCs) are based on the
notion that if you give the equivalent of block grants, or
credits, to countries, they will allocate the resources in
ways that reduce poverty. So while the Bank may not
micromanage or “projectize” the spending, the successes
of PRSCs or other forms of budget support is contingent,
eventually, on the success of government decisions about
how to spend those resources on public health, education
and many other types of programs intended to reduce
poverty and improve the life chances of the poor.
What is missing as an input into the design of most
programs is a genuine evidence base to systematically
support (or refute) that theory of change. Will those
particular investments and institutional changes really
make a positive difference, or do they just sound good?
Have those investments resulted in the desired change
before, in the same country or region, or elsewhere?
We simply have no systematic information about
this outside of a very narrow set of experiences that,
primarily because of historical accident, have been well
evaluated (e.g., conditional cash transfers in Mexico
and Central America).
CGD0502 0527_Engl_6x9.indd 187
8/17/06 2:55:34 PM
188
Rescuing the World Bank
Why So Little Impact Evaluation?
Good impact evaluations are not a core part of the
practice of international development. There are lots of
very good reasons for this:
First, good impact evaluations require a degree
of technical sophistication that is often lacking in the
field of “applied development,” where practitioners are
accustomed to dealing with poor data and unfamiliar
contexts. While many studies compare conditions before
and after a project, such comparisons can be quite
misleading without attention to other factors that might
have also contributed to observed changes. For example,
improvements in population health status might come
from the introduction of new health care services, but
they might also have been induced by rapid economic
growth, migration, personal hygiene, climate change,
or investments in infrastructure. Only by comparing
observed changes among those who benefited from a
project to some other control group is it possible to begin
to disentangle how much of the effects can be attributed
to the project or program itself.
Separating out the changes due to projects from
changes due to other things is a complicated business,
and to date the development community has been
satisfied with weak alternatives, viewing more rigorous
methods as inappropriate to the context of developing
countries. Fortunately, advances in research methods
and increasing capacity around the world to conduct
such impact evaluations is beginning to surmount these
technical difficulties.
Second, demand for the knowledge produced by
impact evaluations tends to be spread out across many
actors and across time. It is only at the moment of
designing a new program that anything can be effectively
done to start an impact evaluation. At that exact
moment, program designers do want the benefit of prior
research, yet have few incentives to invest in starting a
new study. Paradoxically, if they do not invest in a new
study, the same program designers will find themselves
in the same position four or five years later because
the opportunity to learn whether or not the intervention
has an impact was missed.1 Because information from
impact evaluations is a public good, other institutions
and governments that might have learned from the
CGD0502 0527_Engl_6x9.indd 188
8/17/06 2:55:34 PM
189
Selected Essays
experience also lose when these investments in learning
about impact are neglected.
Third, incentives exist at the institutional level to
discourage conducting impact evaluations. Government
agencies involved in social development programs or
international assistance need to generate support from
taxpayers and donors. Because impact evaluations
can go both ways—demonstrating positive or negative
impact—any government or organization that conducts
such research runs the risk of findings that undercut
its ability to raise funds.2 Policymakers and managers
also have more discretion to pick and choose strategic
directions when less is known about what does or does
not work. This can even lead organizations to pressure
researchers to alter, soften or modify unfavorable studies,
or to simply repress the results—despite the fact that
knowledge of what doesn’t work is as useful as learning
what does. Similar disincentives to finding out “bad news”
about program performance exist within institutions like
the World Bank. For task managers, in fact, attempting to
communicate negative results up the managerial “chain
of command” can be one of the least career-savvy moves
one can make.
Fourth, evaluation simply is not the central business of
the Bank, and when material and human resources are
stretched—as they typically are, even in the comparatively
well-endowed environment of the Bank—short-term
operational demands will override the longer-term, more
strategic imperative of evaluation and learning. As one
indication, resources spent to design and implement
impact evaluations were not even recognized as a
separate item in the World Bank’s budgeting system
until this year.
All of these reasons contribute to the situation observed
today. For most types of programs, a body of scientific
evidence about effectiveness is lacking. Bank task
managers designing projects are left to their own devices.
The general strategy is to observe that many other projects
are based on the same theory of change, and on a plethora
of anecdotes, “best practice”-style documentation to
support a given program design, and reference to the
writings of those who are regarded as particularly brilliant
thinkers. This is “eminence-based decision making” rather
than “evidence-based decision making.”
CGD0502 0527_Engl_6x9.indd 189
8/17/06 2:55:35 PM
190
Rescuing the World Bank
A Smarter Future
Fortunately, many have recognized this problem, care
about solving it, and are trying hard to find a way to do so.3
Within the Bank, the IEG advocates for more resources
for good evaluation, and makes heroic efforts to squeeze
knowledge out of the experiences of projects that are
conducted without baseline data, without comparison
groups, sometimes without any impact indicators at all.
In the past couple of years, the World Bank has created
an initiative called the Development IMpact Evaluation
(DIME) Initiative to: increase the number of Bank projects
with impact evaluation components, particularly in
strategic areas and themes; to increase the ability of
staff to design and carry out such evaluations; and
to build a process of systematic learning on effective
development interventions based on lessons learned from
those evaluations.4
The Bank identified five thematic areas to concentrate
its current efforts at impact evaluation: school-based
management and community participation in education;
information for accountability in education; teacher
contracting; conditional cash transfer programs to
improve education outcomes; and slum upgrading
programs. The Bank is also aiming to improve internal
incentives to undertake more systematic development
impact evaluations by explicitly recognizing these studies
as a valued product in their own right.
This represents an important shift in the Bank’s
recognition of the value of impact evaluation—and
particularly in the leadership of key individuals who
have taken on this topic as a personal mission within the
institution. The work of both IEG and the DIME deserve
political and financial support.
But the chances are that this will not be enough.
Even the best intentioned efforts, such as the DIME, will
find it difficult over time to sustain their resources and
maintain enthusiasm and rigor—particularly when some
evaluations will inevitably show that many programs have
been unsuccessful. In most institutions, internal offices
that generate negative reports are subject to pressures to
paint results in a positive light or, over time, find themselves
increasingly isolated and with fewer resources.
CGD0502 0527_Engl_6x9.indd 190
8/17/06 2:55:35 PM
191
Selected Essays
A broader and bolder solution to the problem is required.
Three central elements are required for a lasting and
genuine solution to the problem of lack of knowledge
about what works.
First, we need to use good evaluation methods to get
answers to important questions. This means identifying
the enduring questions, a process that would be done
best if it were done in true partnership between developing
countries and the range of institutions that provide
development finance. The World Bank has made a start by
identifying five thematic areas within its impact evaluation
initiative. But the benefits of concentrating such studies
around enduring questions across agencies and countries
would be even greater. Surely there is an immense
opportunity to learn by collaboration with different
organizations that address similar health, education and
other social problems in profoundly different ways.
Second, we need to use evaluation methods that yield
answers. This means increasing the number of impact
evaluations that use rigorous methods—such as random
assignment and regression discontinuity—and applying
them to a small number of programs from which the
most can be learned. This does not obviate the need
to continue process-oriented evaluation work, which
can be tremendously informative to answer operational
questions, but it does mean there is a new and large
agenda for impact evaluation.
Third, while the overall set of important questions should
be developed by the “interested parties” in development
agencies and developing countries, the impact
evaluations themselves need to be done independently
of the major international agencies and borrowing
country governments. Independent evaluations would
be more credible in the public eye, and less subject to
inappropriate pressures to modify results, interpretations
or presentation. It is still important to work on changing
the culture of the Bank in its entirety and all the myriad
internal incentives to get projects done and implemented
so that evaluation and learning become a regular part of all
the Bank’s activities. But the existence of an independent
source of impact evaluation results—geared to a longer
time frame and toward learning—will avoid many of the
inevitable pressures to restrict the communication of bad
news to higher levels of management.
CGD0502 0527_Engl_6x9.indd 191
8/17/06 2:55:35 PM
192
Rescuing the World Bank
A Proposed Solution
If leaders with vision in a few development agencies
and a few developing country governments put their
minds to it, a major improvement is within grasp. An
international initiative could be established to promote
more and better independent impact evaluations,
undertaking, for example:
• Development of a shared agenda of “enduring
questions” for selective evaluations around which multicountry/multi-agency evaluation could be done.
• Creation and dissemination of standards for
methodological quality of impact evaluation.
• Provision of financial resources for design of impact
evaluations.
• Provision of complementary resources for
implementation of impact evaluations.
• Creation of a registry of impact evaluations.
• Dissemination of impact evaluation results.
• Development of a data clearinghouse to facilitate
reanalysis.
• Support for the development of new and improved
methods.
Appropriately, this would be a collective response to
ensure supply of knowledge, a global public good in the
truest sense. The ideal financing arrangements would be
one based on sharing costs across those who benefit, or
at least those agencies that choose to participate. Those
resources should be additional to the current evaluation
budgets, which have been pared down to subsistence
level. Foundations and other private sector actors who see
the long-term benefits and wish to facilitate generation of
and access to knowledge might also be willing to provide
start-up resources.
The question remains whether the Bank, which uses
so little of its internally produced knowledge from ICRs
and products of the IEG, would develop mechanisms
to apply technical knowledge generated with the input
of an independent facility. The answer to this may
matter relatively little. If such knowledge were part of an
international effort that disseminated findings broadly,
those with whom the Bank works—counterparts who codesign projects with Bank staff—might find themselves
CGD0502 0527_Engl_6x9.indd 192
8/17/06 2:55:35 PM
193
Selected Essays
well equipped with much more evidence about what
works than they have today, and be able to shape the
Bank’s actions in a positive way.
The World Bank’s participation in such an initiative
would be truly win-win.The international community
would benefit from the institution’s expertise and access
to knowledge from the Bank’s tremendous portfolio of
projects.The Bank would benefit from enhanced credibility
and influence, as well as access to knowledge from other
agencies’ projects. Participating actively in the global
process of learning what works is a natural role for the
Bank to take on. Genuine success—making lives better—
depends on it.
Notes
1. Ted O’Donoghue and Matthew Rabin “Doing It Now
or Later,” The American Economic Review 89 (1999):
103–124.
2. Lant Pritchett, “It Pays to Be Ignorant: A Simple
Political Economy of Rigorous Program Evaluation,” The
Journal of Policy Reform 5 (2002): 251–269.
3. This section and those that follow are derived from
discussions of the Evaluation Gap Working Group, and its
final report, “When Will We Ever Learn? Improving Lives
through Improved Impact Evaluation” (Center for Global
Development, 2006). The ideas included here, however,
are those of the essay authors, and not necessarily those
of all working group members. For more information,
see http://www.cgdev.org/section/initatives/_active/
evalgap.
4. Information on DIME from personal communications
with Francois Bourguinon, Paul Gertler and Ariel Fiszbein.
CGD0502 0527_Engl_6x9.indd 193
8/17/06 2:55:35 PM
CGD0502 0527_Engl_6x9.indd 194
8/17/06 2:55:35 PM
The World Bank:
Buy, Sell, or Hold?
by Mark Stoleson
I
t seemed like a simple question, but the World Bank
representative was visibly uncomfortable. Together
with me was a delegation of investors visiting this
small African country to see first-hand the micro-finance
projects we had funded and to identify new development
opportunities. We were visiting the World Bank’s local
offices to learn more about their activities and experience
in the country. The Country Head was explaining how he
planned to allocate his budget of over $100 million and
highlighted one project to build cobble stone roads in
the capital city. Knowing that the World Bank’s mission is
to reduce poverty, I assumed that building cobble stone
roads would lead to a reduction in poverty. So, not being
a development expert and almost thinking aloud, I asked
our host: “How will these roads reduce the country’s
poverty?”
Surprisingly, he struggled to answer the question.
After briefly discussing development theory, he finally
stated that Bank staff on the ground do not have time
to contemplate “ivory tower” notions such as the Bank’s
mission statement or overall goals but rather need to focus
on the day to day business of managing their budgets
and completing projects. But, I asked, if the projects do
not reduce poverty, what is the point?
Only a few weeks before, World Bank President Paul
Wolfowitz had emphasized the need to remedy the Bank’s
historically poor performance in Africa. He told Bank staff
that “in the last 20 years, the number of people living in
extreme poverty in Africa has doubled...in spite of roughly
$200-300 billion in development assistance...[and] it’s
going to be hard to explain ourselves in 5 or 10 years if
that picture remains the same. “In other words, results
matter. Wolfowitz’s comment seemed to imply the need
for the Bank to focus not just on deploying capital and
completing projects, but also on the overall “returns on
Mark Stoleson, Sovereign Global Development
Mark Stoleson is President of Sovereign Global Development. Please see
http://www.sov.com for more information about Sovereign Global.
CGD0502 0527_Engl_6x9.indd 195
195
8/17/06 2:55:35 PM
196
Rescuing the World Bank
investment,” measured by the positive impact on the
lives of people living in extreme poverty. In the office of
the Country Head, however, concepts such as capital
allocation and returns on investment seemed as far away
as Washington DC.
An investor’s view of the world bank
Following our meeting, I couldn’t help but wonder how the
World Bank might look, were it an investment opportunity.
If this were a public company, I wondered, would I buy
its shares? Would the World Bank meet the standards
of performance to which investors such as Sovereign
Global (“Sovereign”) holds its investments?
Given the scale of resources entrusted to the World
Bank, and the importance of its agenda, shouldn’t
someone enquire as to whether that institution is itself
a wise investment? We came to this question with a
perspective shaped both by experience in capital markets
and emerging economies.
Sovereign has been investing in the international capital
markets for over 20 years. During this time, the firm has
provided capital to companies and governments from
Asia and Africa to Latin America and Eastern Europe. Our
investments have spanned many industries ranging from
banking and energy to telecoms, power, and steel.
In every case, we have found that a good investment
has certain characteristics: Firstly, it has a competitive
advantage relative to its market, i.e., it is providing a
product or service more effectively than any alternative
organization. Secondly, it operates under the scrutiny
of independent auditors or evaluators. Finally, it is
accountable to its shareholders who ensure that it follows
a set of coherent strategic goals. In simple terms we
endeavor to measure the organization’s competitiveness,
transparency, and accountability.
Given these core conditions (though not guarantees)
for success, how does the World Bank measure up? We
ran the Bank through a basic set of questions we would
ask when evaluating any investment opportunity.
Competitiveness: Is the Bank competitive at banking?
The Bank’s core customers—developing countries—
increasingly are able to obtain financing from international
debt capital markets. Still, the Bank persists in pursuing
these customers: Over the past five years, 99 percent of
its funds were loaned to countries that have investment
CGD0502 0527_Engl_6x9.indd 196
8/17/06 2:55:36 PM
197
Selected Essays
grade or high-yield bond ratings.2 Yet this core customer
base turns to the Bank for only 1 percent of its total debt
financing needs.3 Any company that has a 1% market
share amongst its target customer base has a dubious
commercial rationale and future prospects.
You would think that given this loss of competitive
standing, the Bank would shift more resources to an
underserved marketplace, such as poor countries whose
prospects for repayment are less certain. In fact, the share
of such loans has gone from 40 percent of available funds
in 1993 to 1 percent more recently.4 What that suggests
to us is that the Bank is chasing customers who do not
want or need Bank funding, while increasingly ignoring
the needs of countries that do.
In addition, the Bank’s lending operations are
unprofitable. Over the last 12 years the Bank has
accumulated net losses of over $3 billion from its lending
operations5—which could have supplemented the Bank’s
own borrowing and investment returns and provided
the Bank with funds to make grants and concessional
loans to the world’s poorest countries. If the Bank were
a better “bank” it would have more funds available for
those who need it the most, but the opposite appears
to be the case.
Is the Bank competitive at reducing poverty? The
Bank’s mission is “to fight poverty with passion and
professionalism for lasting results.” Presumably the
mission is not only to fight, but also to win. After 50
years and $570 billion spent or lent, however, there is no
conclusive data that demonstrates that the World Bank
has made a meaningful impact on its primary mission.
Poverty has declined in East and South Asia—but that
is where World Bank development ideas and lending
have been relatively small. Meanwhile, as highlighted by
Adam Lerrick in recent testimony to the United States
Senate, “The living standards of the poorest nations have
stagnated and even declined as much as 25%.”6 The
Bank has tremendous human and intellectual capital in
the form of its experienced and committed staff. Why
then has the Bank seemingly failed to deliver on its core
mission of reducing poverty?
Transparency: Is the Bank a model of good governance?
The Bank granted or lent $20 billion in 2005—not a huge
amount by international banking standards, but certainly
more than your average regional commercial bank. Yet
CGD0502 0527_Engl_6x9.indd 197
8/17/06 2:55:36 PM
198
Rescuing the World Bank
it does not have a truly independent audit committee or
employ outside auditors to objectively evaluate project
performance. In the post-Enron, Sarbanes-Oxley world,
it is incredible that a multi-billion dollar institution backed
by taxpayer money is operating without independent
oversight. If it were publicly traded on a U.S. stock
exchange, the Bank’s lack of an audit committee alone
would cause it to be de-listed. Largely because of this
blind spot, it is impossible for those either inside or
outside the Bank to know whether it is effective or not,
simply because it lacks any credible or objective metrics
related to performance. Without measurable results there
can be no accountability.
Accountability: Are the Bank’s shareholders aligned in
pursuing a sensible reform agenda? Shareholders with a
common purpose can bring about governance reforms at
under-performing companies. Reform will fail, however,
if certain interests divide shareholders and conquer their
common resolve. Or worse yet, if shareholders simply
do not care about the returns on their investment,
management will never improve performance. The World
Bank does not lack its share of critics, both internal and
external; many of those critics are the Bank’s largest
funders. Yet many of the Bank’s financial backers either do
not agree on a common agenda for reforming the Bank, or
do not care to reform it in the first place. Either way, this
flaw represents an abrogation of fiduciary responsibility
by the Bank’s shareholders to the Bank, to the taxpayers
whose funds support the Bank’s functions, and most of
all, to the poor.
A solution from the private sector
These weaknesses would represent significant reasons
not to invest in the World Bank. Its management and
staff have an impressive and undeniable record of service
and commitment to the cause of development, but the
multiple and conflicting objectives of the Bank and
its shareholders, combined with a total failure at selfgovernance and minimal involvement by shareholders,
have made it impossible for positive “fundamentals”
to emerge.
Any efforts to improve the Bank’s competitive
advantage must start with the creation of independent
governance mechanisms and objective measures of
success and failure. Without these measures, how would
CGD0502 0527_Engl_6x9.indd 198
8/17/06 2:55:36 PM
199
Selected Essays
Bank management and its shareholders discern what
impact, if any, the Bank is really making?7
To make serious efforts toward upgrading its
transparency and accountability, the World Bank, would
benefit from adopting certain measures commonly used
by public companies. Specifically, the World Bank could
consider the merits of:
1. Establishing an Independent Audit Committee. Any
effective audit committee is comprised of members
that are truly independent. That means committee
members could not accept any consulting, advisory, or
other compensatory fee from the Bank, or be affiliated
with the Bank or any of its entities other than as a
member of the Board.
2. Giving the Audit Committee Authority. Independent
audit committees generally report directly to the Board
and have the authority and budget to hire their own
counsel and consultants if necessary.
3. Engaging Top Outside Auditors. Audit committees
are responsible for engaging and managing outside
auditors. A World Bank audit committee could start
by putting out for bid a multi-year, multi-million dollar
auditing engagement that would draw in a qualified
firm to audit the Bank’s financial performance and
projects while establishing proper internal controls so
that the Bank’s management team is equipped with
the information they need to make sound decisions.
The costs of a rigorous audit would be far outweighed
by the benefits of transparency and accountability in
addition to the valuable information the Bank would
be able to provide to its management and the
development industry as a whole about what works
and what does not.
This proposal would offer many benefits. First, it would
harness the inherent power of free market principles in
fashioning new ways to evaluate development projects.
Second, it would transform the World Bank from a
laggard to a leader in accountability and transparency
in the development community. And finally and most
importantly, as the Bank responds to objective and
accurate data about its performance, it will be able to
make adjustments in its business model and operations
CGD0502 0527_Engl_6x9.indd 199
8/17/06 2:55:36 PM
200
Rescuing the World Bank
so that it can more effectively deliver capital to the world’s
poorest communities.
For example, if the Bank were to recognize that larger
and more focused grants have a greater likelihood of
creating transformative change in a region, it would be
able to restructure its grant-making to focus on such
projects. If the Bank saw that its efforts must be combined
within a larger coalition of NGO-led stakeholders on
the ground, then that would be adopted as a preferred
model. If the Bank saw that loans are still a credible
instrument of development aid in certain cases, it could
focus on that mechanism where appropriate and use
grants elsewhere.
Moreover, by establishing credible accountability and
governance measures, the Bank would finally address
the weaknesses that have led to multi-billion dollar losses
in loans. That would assure the Bank’s shareholders—
taxpayers from productive and wealthy nations—that
the cause of development is “worth it” and would finally
give the Bank the unified support and involvement of
its shareholders. Performance, accountability and
transparency are no longer optional in either the public
or private spheres.
We know from experience that the most successful
organizations constantly hold their performance up for
analysis and criticism, and correct those problems as
they emerge. Although there may be some resistance
to an external audit of the World Bank, at the end of the
day full transparency and accountability regarding its
performance will free the Bank from its past, restore its
credibility and relevance, and set it on a new course to
more effectively change the lives of the world’s poor. The
Bank’s mission is too important and its budget too large
to accept anything less.
CGD0502 0527_Engl_6x9.indd 200
8/17/06 2:55:36 PM
201
Selected Essays
Notes
1. World Bank Town Hall Meeting Transcript,
February 6, 2006.
2. See Adam Lerrick, “Has the World Bank Lost
Control?” in this volume.
3. Id.
4. Id.
5. Id.
6. Adam Lerrick, “Is the World Bank’s Word Good
Enough?” Statement Presented to the Committee
on Foreign Relations of the United States Senate,
March 28, 2006.
7. See Ruth Levine and William D. Savedoff, “The
Evaluation Agenda,” in this volume.
CGD0502 0527_Engl_6x9.indd 201
8/17/06 2:55:36 PM